diff --git a/.github/workflows/master.yml b/.github/workflows/master.yml index e662a5b6f98..05654926fd7 100644 --- a/.github/workflows/master.yml +++ b/.github/workflows/master.yml @@ -205,6 +205,12 @@ jobs: with: build_name: binary_amd64_compat checkout_depth: 0 + BuilderBinAmd64Musl: + needs: [DockerHubPush] + uses: ./.github/workflows/reusable_build.yml + with: + build_name: binary_amd64_musl + checkout_depth: 0 BuilderBinAarch64V80Compat: needs: [DockerHubPush] uses: ./.github/workflows/reusable_build.yml diff --git a/.github/workflows/pull_request.yml b/.github/workflows/pull_request.yml index c92fa52ed4e..a6631a93766 100644 --- a/.github/workflows/pull_request.yml +++ b/.github/workflows/pull_request.yml @@ -242,6 +242,11 @@ jobs: uses: ./.github/workflows/reusable_build.yml with: build_name: binary_amd64_compat + BuilderBinAmd64Musl: + needs: [FastTest, StyleCheck] + uses: ./.github/workflows/reusable_build.yml + with: + build_name: binary_amd64_musl BuilderBinAarch64V80Compat: needs: [FastTest, StyleCheck] uses: ./.github/workflows/reusable_build.yml diff --git a/base/glibc-compatibility/glibc-compatibility.c b/base/glibc-compatibility/glibc-compatibility.c index 49bb81a58be..738cda47877 100644 --- a/base/glibc-compatibility/glibc-compatibility.c +++ b/base/glibc-compatibility/glibc-compatibility.c @@ -30,7 +30,6 @@ int __gai_sigqueue(int sig, const union sigval val, pid_t caller_pid) } -#include #include #include diff --git a/base/poco/Foundation/CMakeLists.txt b/base/poco/Foundation/CMakeLists.txt index d0dde8a51a5..dfb41a33fb1 100644 --- a/base/poco/Foundation/CMakeLists.txt +++ b/base/poco/Foundation/CMakeLists.txt @@ -55,7 +55,6 @@ set (SRCS src/DigestStream.cpp src/DirectoryIterator.cpp src/DirectoryIteratorStrategy.cpp - src/DirectoryWatcher.cpp src/Environment.cpp src/Error.cpp src/ErrorHandler.cpp diff --git a/base/poco/Foundation/include/Poco/DirectoryWatcher.h b/base/poco/Foundation/include/Poco/DirectoryWatcher.h deleted file mode 100644 index 00964a5512a..00000000000 --- a/base/poco/Foundation/include/Poco/DirectoryWatcher.h +++ /dev/null @@ -1,228 +0,0 @@ -// -// DirectoryWatcher.h -// -// Library: Foundation -// Package: Filesystem -// Module: DirectoryWatcher -// -// Definition of the DirectoryWatcher class. -// -// Copyright (c) 2012, Applied Informatics Software Engineering GmbH. -// and Contributors. -// -// SPDX-License-Identifier: BSL-1.0 -// - - -#ifndef Foundation_DirectoryWatcher_INCLUDED -#define Foundation_DirectoryWatcher_INCLUDED - - -#include "Poco/Foundation.h" - - -#ifndef POCO_NO_INOTIFY - - -# include "Poco/AtomicCounter.h" -# include "Poco/BasicEvent.h" -# include "Poco/File.h" -# include "Poco/Runnable.h" -# include "Poco/Thread.h" - - -namespace Poco -{ - - -class DirectoryWatcherStrategy; - - -class Foundation_API DirectoryWatcher : protected Runnable -/// This class is used to get notifications about changes -/// to the filesystem, more specifically, to a specific -/// directory. Changes to a directory are reported via -/// events. -/// -/// A thread will be created that watches the specified -/// directory for changes. Events are reported in the context -/// of this thread. -/// -/// Note that changes to files in subdirectories of the watched -/// directory are not reported. Separate DirectoryWatcher objects -/// must be created for these directories if they should be watched. -/// -/// Changes to file attributes are not reported. -/// -/// On Windows, this class is implemented using FindFirstChangeNotification()/FindNextChangeNotification(). -/// On Linux, this class is implemented using inotify. -/// On FreeBSD and Darwin (Mac OS X, iOS), this class uses kevent/kqueue. -/// On all other platforms, the watched directory is periodically scanned -/// for changes. This can negatively affect performance if done too often. -/// Therefore, the interval in which scans are done can be specified in -/// the constructor. Note that periodic scanning will also be done on FreeBSD -/// and Darwin if events for changes to files (DW_ITEM_MODIFIED) are enabled. -/// -/// DW_ITEM_MOVED_FROM and DW_ITEM_MOVED_TO events will only be reported -/// on Linux. On other platforms, a file rename or move operation -/// will be reported via a DW_ITEM_REMOVED and a DW_ITEM_ADDED event. -/// The order of these two events is not defined. -/// -/// An event mask can be specified to enable only certain events. -{ -public: - enum DirectoryEventType - { - DW_ITEM_ADDED = 1, - /// A new item has been created and added to the directory. - - DW_ITEM_REMOVED = 2, - /// An item has been removed from the directory. - - DW_ITEM_MODIFIED = 4, - /// An item has been modified. - - DW_ITEM_MOVED_FROM = 8, - /// An item has been renamed or moved. This event delivers the old name. - - DW_ITEM_MOVED_TO = 16 - /// An item has been renamed or moved. This event delivers the new name. - }; - - enum DirectoryEventMask - { - DW_FILTER_ENABLE_ALL = 31, - /// Enables all event types. - - DW_FILTER_DISABLE_ALL = 0 - /// Disables all event types. - }; - - enum - { - DW_DEFAULT_SCAN_INTERVAL = 5 /// Default scan interval for platforms that don't provide a native notification mechanism. - }; - - struct DirectoryEvent - { - DirectoryEvent(const File & f, DirectoryEventType ev) : item(f), event(ev) { } - - const File & item; /// The directory or file that has been changed. - DirectoryEventType event; /// The kind of event. - }; - - BasicEvent itemAdded; - /// Fired when a file or directory has been created or added to the directory. - - BasicEvent itemRemoved; - /// Fired when a file or directory has been removed from the directory. - - BasicEvent itemModified; - /// Fired when a file or directory has been modified. - - BasicEvent itemMovedFrom; - /// Fired when a file or directory has been renamed. This event delivers the old name. - - BasicEvent itemMovedTo; - /// Fired when a file or directory has been moved. This event delivers the new name. - - BasicEvent scanError; - /// Fired when an error occurs while scanning for changes. - - DirectoryWatcher(const std::string & path, int eventMask = DW_FILTER_ENABLE_ALL, int scanInterval = DW_DEFAULT_SCAN_INTERVAL); - /// Creates a DirectoryWatcher for the directory given in path. - /// To enable only specific events, an eventMask can be specified by - /// OR-ing the desired event IDs (e.g., DW_ITEM_ADDED | DW_ITEM_MODIFIED). - /// On platforms where no native filesystem notifications are available, - /// scanInterval specifies the interval in seconds between scans - /// of the directory. - - DirectoryWatcher(const File & directory, int eventMask = DW_FILTER_ENABLE_ALL, int scanInterval = DW_DEFAULT_SCAN_INTERVAL); - /// Creates a DirectoryWatcher for the specified directory - /// To enable only specific events, an eventMask can be specified by - /// OR-ing the desired event IDs (e.g., DW_ITEM_ADDED | DW_ITEM_MODIFIED). - /// On platforms where no native filesystem notifications are available, - /// scanInterval specifies the interval in seconds between scans - /// of the directory. - - ~DirectoryWatcher(); - /// Destroys the DirectoryWatcher. - - void suspendEvents(); - /// Suspends sending of events. Can be called multiple times, but every - /// call to suspendEvent() must be matched by a call to resumeEvents(). - - void resumeEvents(); - /// Resumes events, after they have been suspended with a call to suspendEvents(). - - bool eventsSuspended() const; - /// Returns true iff events are suspended. - - int eventMask() const; - /// Returns the value of the eventMask passed to the constructor. - - int scanInterval() const; - /// Returns the scan interval in seconds. - - const File & directory() const; - /// Returns the directory being watched. - - bool supportsMoveEvents() const; - /// Returns true iff the platform supports DW_ITEM_MOVED_FROM/itemMovedFrom and - /// DW_ITEM_MOVED_TO/itemMovedTo events. - -protected: - void init(); - void stop(); - void run(); - -private: - DirectoryWatcher(); - DirectoryWatcher(const DirectoryWatcher &); - DirectoryWatcher & operator=(const DirectoryWatcher &); - - Thread _thread; - File _directory; - int _eventMask; - AtomicCounter _eventsSuspended; - int _scanInterval; - DirectoryWatcherStrategy * _pStrategy; -}; - - -// -// inlines -// - - -inline bool DirectoryWatcher::eventsSuspended() const -{ - return _eventsSuspended.value() > 0; -} - - -inline int DirectoryWatcher::eventMask() const -{ - return _eventMask; -} - - -inline int DirectoryWatcher::scanInterval() const -{ - return _scanInterval; -} - - -inline const File & DirectoryWatcher::directory() const -{ - return _directory; -} - - -} // namespace Poco - - -#endif // POCO_NO_INOTIFY - - -#endif // Foundation_DirectoryWatcher_INCLUDED diff --git a/base/poco/Foundation/src/DirectoryWatcher.cpp b/base/poco/Foundation/src/DirectoryWatcher.cpp deleted file mode 100644 index b559da65e09..00000000000 --- a/base/poco/Foundation/src/DirectoryWatcher.cpp +++ /dev/null @@ -1,602 +0,0 @@ -// -// DirectoryWatcher.cpp -// -// Library: Foundation -// Package: Filesystem -// Module: DirectoryWatcher -// -// Copyright (c) 2012, Applied Informatics Software Engineering GmbH. -// and Contributors. -// -// SPDX-License-Identifier: BSL-1.0 -// - - -#include "Poco/DirectoryWatcher.h" - - -#ifndef POCO_NO_INOTIFY - - -#include "Poco/Path.h" -#include "Poco/Glob.h" -#include "Poco/DirectoryIterator.h" -#include "Poco/Event.h" -#include "Poco/Exception.h" -#include "Poco/Buffer.h" -#if POCO_OS == POCO_OS_LINUX || POCO_OS == POCO_OS_ANDROID - #include - #include - #include -#elif POCO_OS == POCO_OS_MAC_OS_X || POCO_OS == POCO_OS_FREE_BSD - #include - #include - #include - #include - #include - #if (POCO_OS == POCO_OS_FREE_BSD) && !defined(O_EVTONLY) - #define O_EVTONLY 0x8000 - #endif -#endif -#include -#include -#include - -namespace Poco { - - -class DirectoryWatcherStrategy -{ -public: - DirectoryWatcherStrategy(DirectoryWatcher& owner): - _owner(owner) - { - } - - virtual ~DirectoryWatcherStrategy() - { - } - - DirectoryWatcher& owner() - { - return _owner; - } - - virtual void run() = 0; - virtual void stop() = 0; - virtual bool supportsMoveEvents() const = 0; - -protected: - struct ItemInfo - { - ItemInfo(): - size(0) - { - } - - ItemInfo(const ItemInfo& other): - path(other.path), - size(other.size), - lastModified(other.lastModified) - { - } - - explicit ItemInfo(const File& f): - path(f.path()), - size(f.isFile() ? f.getSize() : 0), - lastModified(f.getLastModified()) - { - } - - std::string path; - File::FileSize size; - Timestamp lastModified; - }; - typedef std::map ItemInfoMap; - - void scan(ItemInfoMap& entries) - { - DirectoryIterator it(owner().directory()); - DirectoryIterator end; - while (it != end) - { - entries[it.path().getFileName()] = ItemInfo(*it); - ++it; - } - } - - void compare(ItemInfoMap& oldEntries, ItemInfoMap& newEntries) - { - for (ItemInfoMap::iterator itn = newEntries.begin(); itn != newEntries.end(); ++itn) - { - ItemInfoMap::iterator ito = oldEntries.find(itn->first); - if (ito != oldEntries.end()) - { - if ((owner().eventMask() & DirectoryWatcher::DW_ITEM_MODIFIED) && !owner().eventsSuspended()) - { - if (itn->second.size != ito->second.size || itn->second.lastModified != ito->second.lastModified) - { - Poco::File f(itn->second.path); - DirectoryWatcher::DirectoryEvent ev(f, DirectoryWatcher::DW_ITEM_MODIFIED); - owner().itemModified(&owner(), ev); - } - } - oldEntries.erase(ito); - } - else if ((owner().eventMask() & DirectoryWatcher::DW_ITEM_ADDED) && !owner().eventsSuspended()) - { - Poco::File f(itn->second.path); - DirectoryWatcher::DirectoryEvent ev(f, DirectoryWatcher::DW_ITEM_ADDED); - owner().itemAdded(&owner(), ev); - } - } - if ((owner().eventMask() & DirectoryWatcher::DW_ITEM_REMOVED) && !owner().eventsSuspended()) - { - for (ItemInfoMap::iterator it = oldEntries.begin(); it != oldEntries.end(); ++it) - { - Poco::File f(it->second.path); - DirectoryWatcher::DirectoryEvent ev(f, DirectoryWatcher::DW_ITEM_REMOVED); - owner().itemRemoved(&owner(), ev); - } - } - } - -private: - DirectoryWatcherStrategy(); - DirectoryWatcherStrategy(const DirectoryWatcherStrategy&); - DirectoryWatcherStrategy& operator = (const DirectoryWatcherStrategy&); - - DirectoryWatcher& _owner; -}; - - -#if POCO_OS == POCO_OS_WINDOWS_NT - - -class WindowsDirectoryWatcherStrategy: public DirectoryWatcherStrategy -{ -public: - WindowsDirectoryWatcherStrategy(DirectoryWatcher& owner): - DirectoryWatcherStrategy(owner) - { - _hStopped = CreateEventW(NULL, FALSE, FALSE, NULL); - if (!_hStopped) - throw SystemException("cannot create event"); - } - - ~WindowsDirectoryWatcherStrategy() - { - CloseHandle(_hStopped); - } - - void run() - { - ItemInfoMap entries; - scan(entries); - - DWORD filter = FILE_NOTIFY_CHANGE_FILE_NAME | FILE_NOTIFY_CHANGE_DIR_NAME; - if (owner().eventMask() & DirectoryWatcher::DW_ITEM_MODIFIED) - filter |= FILE_NOTIFY_CHANGE_SIZE | FILE_NOTIFY_CHANGE_LAST_WRITE; - - std::string path(owner().directory().path()); - HANDLE hChange = FindFirstChangeNotificationA(path.c_str(), FALSE, filter); - - if (hChange == INVALID_HANDLE_VALUE) - { - try - { - FileImpl::handleLastErrorImpl(path); - } - catch (Poco::Exception& exc) - { - owner().scanError(&owner(), exc); - } - return; - } - - bool stopped = false; - while (!stopped) - { - try - { - HANDLE h[2]; - h[0] = _hStopped; - h[1] = hChange; - switch (WaitForMultipleObjects(2, h, FALSE, INFINITE)) - { - case WAIT_OBJECT_0: - stopped = true; - break; - case WAIT_OBJECT_0 + 1: - { - ItemInfoMap newEntries; - scan(newEntries); - compare(entries, newEntries); - std::swap(entries, newEntries); - if (FindNextChangeNotification(hChange) == FALSE) - { - FileImpl::handleLastErrorImpl(path); - } - } - break; - default: - throw SystemException("failed to wait for directory changes"); - } - } - catch (Poco::Exception& exc) - { - owner().scanError(&owner(), exc); - } - } - FindCloseChangeNotification(hChange); - } - - void stop() - { - SetEvent(_hStopped); - } - - bool supportsMoveEvents() const - { - return false; - } - -private: - HANDLE _hStopped; -}; - - -#elif POCO_OS == POCO_OS_LINUX || POCO_OS == POCO_OS_ANDROID - - -class LinuxDirectoryWatcherStrategy: public DirectoryWatcherStrategy -{ -public: - LinuxDirectoryWatcherStrategy(DirectoryWatcher& owner): - DirectoryWatcherStrategy(owner), - _fd(-1), - _stopped(false) - { - _fd = inotify_init(); - if (_fd == -1) throw Poco::IOException("cannot initialize inotify", errno); - } - - ~LinuxDirectoryWatcherStrategy() - { - close(_fd); - } - - void run() - { - int mask = 0; - if (owner().eventMask() & DirectoryWatcher::DW_ITEM_ADDED) - mask |= IN_CREATE; - if (owner().eventMask() & DirectoryWatcher::DW_ITEM_REMOVED) - mask |= IN_DELETE; - if (owner().eventMask() & DirectoryWatcher::DW_ITEM_MODIFIED) - mask |= IN_MODIFY; - if (owner().eventMask() & DirectoryWatcher::DW_ITEM_MOVED_FROM) - mask |= IN_MOVED_FROM; - if (owner().eventMask() & DirectoryWatcher::DW_ITEM_MOVED_TO) - mask |= IN_MOVED_TO; - int wd = inotify_add_watch(_fd, owner().directory().path().c_str(), mask); - if (wd == -1) - { - try - { - FileImpl::handleLastErrorImpl(owner().directory().path()); - } - catch (Poco::Exception& exc) - { - owner().scanError(&owner(), exc); - } - } - - Poco::Buffer buffer(4096); - while (!_stopped.load(std::memory_order_relaxed)) - { - fd_set fds; - FD_ZERO(&fds); - FD_SET(_fd, &fds); - - struct timeval tv; - tv.tv_sec = 0; - tv.tv_usec = 200000; - - if (select(_fd + 1, &fds, NULL, NULL, &tv) == 1) - { - int n = read(_fd, buffer.begin(), buffer.size()); - int i = 0; - if (n > 0) - { - while (n > 0) - { - struct inotify_event* event = reinterpret_cast(buffer.begin() + i); - - if (event->len > 0) - { - if (!owner().eventsSuspended()) - { - Poco::Path p(owner().directory().path()); - p.makeDirectory(); - p.setFileName(event->name); - Poco::File f(p.toString()); - - if ((event->mask & IN_CREATE) && (owner().eventMask() & DirectoryWatcher::DW_ITEM_ADDED)) - { - DirectoryWatcher::DirectoryEvent ev(f, DirectoryWatcher::DW_ITEM_ADDED); - owner().itemAdded(&owner(), ev); - } - if ((event->mask & IN_DELETE) && (owner().eventMask() & DirectoryWatcher::DW_ITEM_REMOVED)) - { - DirectoryWatcher::DirectoryEvent ev(f, DirectoryWatcher::DW_ITEM_REMOVED); - owner().itemRemoved(&owner(), ev); - } - if ((event->mask & IN_MODIFY) && (owner().eventMask() & DirectoryWatcher::DW_ITEM_MODIFIED)) - { - DirectoryWatcher::DirectoryEvent ev(f, DirectoryWatcher::DW_ITEM_MODIFIED); - owner().itemModified(&owner(), ev); - } - if ((event->mask & IN_MOVED_FROM) && (owner().eventMask() & DirectoryWatcher::DW_ITEM_MOVED_FROM)) - { - DirectoryWatcher::DirectoryEvent ev(f, DirectoryWatcher::DW_ITEM_MOVED_FROM); - owner().itemMovedFrom(&owner(), ev); - } - if ((event->mask & IN_MOVED_TO) && (owner().eventMask() & DirectoryWatcher::DW_ITEM_MOVED_TO)) - { - DirectoryWatcher::DirectoryEvent ev(f, DirectoryWatcher::DW_ITEM_MOVED_TO); - owner().itemMovedTo(&owner(), ev); - } - } - } - - i += sizeof(inotify_event) + event->len; - n -= sizeof(inotify_event) + event->len; - } - } - } - } - } - - void stop() - { - _stopped.store(true, std::memory_order_relaxed); - } - - bool supportsMoveEvents() const - { - return true; - } - -private: - int _fd; - std::atomic _stopped; -}; - - -#elif POCO_OS == POCO_OS_MAC_OS_X || POCO_OS == POCO_OS_FREE_BSD - - -class BSDDirectoryWatcherStrategy: public DirectoryWatcherStrategy -{ -public: - BSDDirectoryWatcherStrategy(DirectoryWatcher& owner): - DirectoryWatcherStrategy(owner), - _queueFD(-1), - _dirFD(-1), - _stopped(false) - { - _dirFD = open(owner.directory().path().c_str(), O_EVTONLY); - if (_dirFD < 0) throw Poco::FileNotFoundException(owner.directory().path()); - _queueFD = kqueue(); - if (_queueFD < 0) - { - close(_dirFD); - throw Poco::SystemException("Cannot create kqueue", errno); - } - } - - ~BSDDirectoryWatcherStrategy() - { - close(_dirFD); - close(_queueFD); - } - - void run() - { - Poco::Timestamp lastScan; - ItemInfoMap entries; - scan(entries); - - while (!_stopped.load(std::memory_order_relaxed)) - { - struct timespec timeout; - timeout.tv_sec = 0; - timeout.tv_nsec = 200000000; - unsigned eventFilter = NOTE_WRITE; - struct kevent event; - struct kevent eventData; - EV_SET(&event, _dirFD, EVFILT_VNODE, EV_ADD | EV_CLEAR, eventFilter, 0, 0); - int nEvents = kevent(_queueFD, &event, 1, &eventData, 1, &timeout); - if (nEvents < 0 || eventData.flags == EV_ERROR) - { - try - { - FileImpl::handleLastErrorImpl(owner().directory().path()); - } - catch (Poco::Exception& exc) - { - owner().scanError(&owner(), exc); - } - } - else if (nEvents > 0 || ((owner().eventMask() & DirectoryWatcher::DW_ITEM_MODIFIED) && lastScan.isElapsed(owner().scanInterval()*1000000))) - { - ItemInfoMap newEntries; - scan(newEntries); - compare(entries, newEntries); - std::swap(entries, newEntries); - lastScan.update(); - } - } - } - - void stop() - { - _stopped.store(true, std::memory_order_relaxed); - } - - bool supportsMoveEvents() const - { - return false; - } - -private: - int _queueFD; - int _dirFD; - std::atomic _stopped; -}; - - -#else - - -class PollingDirectoryWatcherStrategy: public DirectoryWatcherStrategy -{ -public: - PollingDirectoryWatcherStrategy(DirectoryWatcher& owner): - DirectoryWatcherStrategy(owner) - { - } - - ~PollingDirectoryWatcherStrategy() - { - } - - void run() - { - ItemInfoMap entries; - scan(entries); - while (!_stopped.tryWait(1000*owner().scanInterval())) - { - try - { - ItemInfoMap newEntries; - scan(newEntries); - compare(entries, newEntries); - std::swap(entries, newEntries); - } - catch (Poco::Exception& exc) - { - owner().scanError(&owner(), exc); - } - } - } - - void stop() - { - _stopped.set(); - } - - bool supportsMoveEvents() const - { - return false; - } - -private: - Poco::Event _stopped; -}; - - -#endif - - -DirectoryWatcher::DirectoryWatcher(const std::string& path, int eventMask, int scanInterval): - _directory(path), - _eventMask(eventMask), - _scanInterval(scanInterval) -{ - init(); -} - - -DirectoryWatcher::DirectoryWatcher(const Poco::File& directory, int eventMask, int scanInterval): - _directory(directory), - _eventMask(eventMask), - _scanInterval(scanInterval) -{ - init(); -} - - -DirectoryWatcher::~DirectoryWatcher() -{ - try - { - stop(); - delete _pStrategy; - } - catch (...) - { - poco_unexpected(); - } -} - - -void DirectoryWatcher::suspendEvents() -{ - poco_assert (_eventsSuspended > 0); - - _eventsSuspended--; -} - - -void DirectoryWatcher::resumeEvents() -{ - _eventsSuspended++; -} - - -void DirectoryWatcher::init() -{ - if (!_directory.exists()) - throw Poco::FileNotFoundException(_directory.path()); - - if (!_directory.isDirectory()) - throw Poco::InvalidArgumentException("not a directory", _directory.path()); - -#if POCO_OS == POCO_OS_WINDOWS_NT - _pStrategy = new WindowsDirectoryWatcherStrategy(*this); -#elif POCO_OS == POCO_OS_LINUX || POCO_OS == POCO_OS_ANDROID - _pStrategy = new LinuxDirectoryWatcherStrategy(*this); -#elif POCO_OS == POCO_OS_MAC_OS_X || POCO_OS == POCO_OS_FREE_BSD - _pStrategy = new BSDDirectoryWatcherStrategy(*this); -#else - _pStrategy = new PollingDirectoryWatcherStrategy(*this); -#endif - _thread.start(*this); -} - - -void DirectoryWatcher::run() -{ - _pStrategy->run(); -} - - -void DirectoryWatcher::stop() -{ - _pStrategy->stop(); - _thread.join(); -} - - -bool DirectoryWatcher::supportsMoveEvents() const -{ - return _pStrategy->supportsMoveEvents(); -} - - -} // namespace Poco - - -#endif // POCO_NO_INOTIFY diff --git a/cmake/target.cmake b/cmake/target.cmake index 887f79bf24e..1680715d15f 100644 --- a/cmake/target.cmake +++ b/cmake/target.cmake @@ -42,10 +42,8 @@ if (CMAKE_CROSSCOMPILING) if (ARCH_AARCH64) # FIXME: broken dependencies set (ENABLE_GRPC OFF CACHE INTERNAL "") - set (ENABLE_SENTRY OFF CACHE INTERNAL "") elseif (ARCH_PPC64LE) set (ENABLE_GRPC OFF CACHE INTERNAL "") - set (ENABLE_SENTRY OFF CACHE INTERNAL "") elseif (ARCH_RISCV64) # RISC-V support is preliminary set (GLIBC_COMPATIBILITY OFF CACHE INTERNAL "") @@ -73,19 +71,10 @@ if (CMAKE_CROSSCOMPILING) message (FATAL_ERROR "Trying to cross-compile to unsupported system: ${CMAKE_SYSTEM_NAME}!") endif () - if (USE_MUSL) - # use of undeclared identifier 'PTHREAD_RECURSIVE_MUTEX_INITIALIZER_NP' - set (ENABLE_SENTRY OFF CACHE INTERNAL "") - set (ENABLE_ODBC OFF CACHE INTERNAL "") - set (ENABLE_GRPC OFF CACHE INTERNAL "") - set (ENABLE_HDFS OFF CACHE INTERNAL "") - set (ENABLE_EMBEDDED_COMPILER OFF CACHE INTERNAL "") - # use of drand48_data - set (ENABLE_AZURE_BLOB_STORAGE OFF CACHE INTERNAL "") - endif () - - # Don't know why but CXX_STANDARD doesn't work for cross-compilation - set (CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -std=c++20") - message (STATUS "Cross-compiling for target: ${CMAKE_CXX_COMPILE_TARGET}") endif () + +if (USE_MUSL) + # Does not work for unknown reason + set (ENABLE_RUST OFF CACHE INTERNAL "") +endif () diff --git a/contrib/CMakeLists.txt b/contrib/CMakeLists.txt index 37b39d4b208..e91ab38ca00 100644 --- a/contrib/CMakeLists.txt +++ b/contrib/CMakeLists.txt @@ -134,9 +134,9 @@ add_contrib (libuv-cmake libuv) add_contrib (liburing-cmake liburing) add_contrib (amqpcpp-cmake AMQP-CPP) # requires: libuv add_contrib (cassandra-cmake cassandra) # requires: libuv +add_contrib (curl-cmake curl) +add_contrib (azure-cmake azure) # requires: curl if (NOT OS_DARWIN) - add_contrib (curl-cmake curl) - add_contrib (azure-cmake azure) # requires: curl add_contrib (sentry-native-cmake sentry-native) # requires: curl endif() add_contrib (fmtlib-cmake fmtlib) diff --git a/contrib/azure b/contrib/azure index 096049bf24f..352ff0a61cb 160000 --- a/contrib/azure +++ b/contrib/azure @@ -1 +1 @@ -Subproject commit 096049bf24fffafcaccc132b9367694532716731 +Subproject commit 352ff0a61cb319ac1cc38c4058443ddf70147530 diff --git a/contrib/curl-cmake/CMakeLists.txt b/contrib/curl-cmake/CMakeLists.txt index 7e86352befc..e74629e57b3 100644 --- a/contrib/curl-cmake/CMakeLists.txt +++ b/contrib/curl-cmake/CMakeLists.txt @@ -10,7 +10,7 @@ set (LIBRARY_DIR "${ClickHouse_SOURCE_DIR}/contrib/curl") set (SRCS "${LIBRARY_DIR}/lib/altsvc.c" "${LIBRARY_DIR}/lib/amigaos.c" - "${LIBRARY_DIR}/lib/asyn-thread.c" + "${LIBRARY_DIR}/lib/asyn-ares.c" "${LIBRARY_DIR}/lib/base64.c" "${LIBRARY_DIR}/lib/bufq.c" "${LIBRARY_DIR}/lib/bufref.c" @@ -165,13 +165,14 @@ target_compile_definitions (_curl PRIVATE libcurl_EXPORTS OS="${CMAKE_SYSTEM_NAME}" ) + target_include_directories (_curl SYSTEM PUBLIC "${LIBRARY_DIR}/include" "${LIBRARY_DIR}/lib" . # curl_config.h ) -target_link_libraries (_curl PRIVATE OpenSSL::SSL) +target_link_libraries (_curl PRIVATE OpenSSL::SSL ch_contrib::c-ares) # The library is large - avoid bloat (XXX: is it?) if (OMIT_HEAVY_DEBUG_SYMBOLS) diff --git a/contrib/curl-cmake/curl_config.h b/contrib/curl-cmake/curl_config.h index f56ba3eccd5..a38aa60fe6d 100644 --- a/contrib/curl-cmake/curl_config.h +++ b/contrib/curl-cmake/curl_config.h @@ -50,3 +50,4 @@ #define ENABLE_IPV6 #define USE_OPENSSL #define USE_THREADS_POSIX +#define USE_ARES diff --git a/contrib/libhdfs3 b/contrib/libhdfs3 index bdcb91354b1..b9598e60167 160000 --- a/contrib/libhdfs3 +++ b/contrib/libhdfs3 @@ -1 +1 @@ -Subproject commit bdcb91354b1c05b21e73043a112a6f1e3b013497 +Subproject commit b9598e6016720a7c088bfe85ce1fa0410f9d2103 diff --git a/contrib/libhdfs3-cmake/CMakeLists.txt b/contrib/libhdfs3-cmake/CMakeLists.txt index 4278575fd7f..8cd951af746 100644 --- a/contrib/libhdfs3-cmake/CMakeLists.txt +++ b/contrib/libhdfs3-cmake/CMakeLists.txt @@ -26,6 +26,11 @@ ADD_DEFINITIONS(-D__STDC_FORMAT_MACROS) ADD_DEFINITIONS(-D_GNU_SOURCE) ADD_DEFINITIONS(-D_GLIBCXX_USE_NANOSLEEP) ADD_DEFINITIONS(-DHAVE_NANOSLEEP) + +if (USE_MUSL) + ADD_DEFINITIONS(-DSTRERROR_R_RETURN_INT) +endif () + set(HAVE_STEADY_CLOCK 1) set(HAVE_NESTED_EXCEPTION 1) SET(HAVE_BOOST_CHRONO 0) diff --git a/contrib/libxml2-cmake/linux_x86_64/include/libxml/xmlversion.h b/contrib/libxml2-cmake/linux_x86_64/include/libxml/xmlversion.h index 9eabfaa50c8..c2faeb47cb1 100644 --- a/contrib/libxml2-cmake/linux_x86_64/include/libxml/xmlversion.h +++ b/contrib/libxml2-cmake/linux_x86_64/include/libxml/xmlversion.h @@ -270,7 +270,7 @@ XMLPUBFUN void XMLCALL xmlCheckVersion(int version); * * Whether iconv support is available */ -#if 1 +#if 0 #define LIBXML_ICONV_ENABLED #endif @@ -499,5 +499,3 @@ XMLPUBFUN void XMLCALL xmlCheckVersion(int version); } #endif /* __cplusplus */ #endif - - diff --git a/contrib/llvm-project b/contrib/llvm-project index e7b8befca85..1834e42289c 160000 --- a/contrib/llvm-project +++ b/contrib/llvm-project @@ -1 +1 @@ -Subproject commit e7b8befca85c8b847614432dba250c22d35fbae0 +Subproject commit 1834e42289c58402c804a87be4d489892b88f3ec diff --git a/contrib/rocksdb-cmake/CMakeLists.txt b/contrib/rocksdb-cmake/CMakeLists.txt index 7d7666dff87..c4220ba90ac 100644 --- a/contrib/rocksdb-cmake/CMakeLists.txt +++ b/contrib/rocksdb-cmake/CMakeLists.txt @@ -117,7 +117,7 @@ endif() add_definitions(-DROCKSDB_PLATFORM_POSIX -DROCKSDB_LIB_IO_POSIX) -if (OS_LINUX OR OS_FREEBSD) +if ((OS_LINUX OR OS_FREEBSD) AND NOT USE_MUSL) add_definitions(-DROCKSDB_PTHREAD_ADAPTIVE_MUTEX) endif() diff --git a/contrib/sentry-native b/contrib/sentry-native index ae10fb8c224..bc359f86cbf 160000 --- a/contrib/sentry-native +++ b/contrib/sentry-native @@ -1 +1 @@ -Subproject commit ae10fb8c224c3f41571446e1ed7fd57b9e5e366b +Subproject commit bc359f86cbf0f73f6fd4b6bfb4ede0c1f8c9400f diff --git a/contrib/sentry-native-cmake/CMakeLists.txt b/contrib/sentry-native-cmake/CMakeLists.txt index 377f955f856..6364e75db28 100644 --- a/contrib/sentry-native-cmake/CMakeLists.txt +++ b/contrib/sentry-native-cmake/CMakeLists.txt @@ -13,6 +13,7 @@ set (SRC_DIR "${ClickHouse_SOURCE_DIR}/contrib/sentry-native") set (SRCS ${SRC_DIR}/vendor/mpack.c + ${SRC_DIR}/vendor/stb_sprintf.c ${SRC_DIR}/src/sentry_alloc.c ${SRC_DIR}/src/sentry_backend.c ${SRC_DIR}/src/sentry_core.c @@ -21,6 +22,7 @@ set (SRCS ${SRC_DIR}/src/sentry_json.c ${SRC_DIR}/src/sentry_logger.c ${SRC_DIR}/src/sentry_options.c + ${SRC_DIR}/src/sentry_os.c ${SRC_DIR}/src/sentry_random.c ${SRC_DIR}/src/sentry_ratelimiter.c ${SRC_DIR}/src/sentry_scope.c @@ -29,6 +31,7 @@ set (SRCS ${SRC_DIR}/src/sentry_string.c ${SRC_DIR}/src/sentry_sync.c ${SRC_DIR}/src/sentry_transport.c + ${SRC_DIR}/src/sentry_tracing.c ${SRC_DIR}/src/sentry_utils.c ${SRC_DIR}/src/sentry_uuid.c ${SRC_DIR}/src/sentry_value.c diff --git a/contrib/unixodbc-cmake/CMakeLists.txt b/contrib/unixodbc-cmake/CMakeLists.txt index 3317654cd67..6fbe8c14ebb 100644 --- a/contrib/unixodbc-cmake/CMakeLists.txt +++ b/contrib/unixodbc-cmake/CMakeLists.txt @@ -1,7 +1,7 @@ option (ENABLE_ODBC "Enable ODBC library" ${ENABLE_LIBRARIES}) -if (NOT OS_LINUX) +if (NOT OS_LINUX OR USE_MUSL) if (ENABLE_ODBC) - message(STATUS "ODBC is only supported on Linux") + message(STATUS "ODBC is only supported on Linux with dynamic linking") endif() set (ENABLE_ODBC OFF CACHE INTERNAL "") endif () diff --git a/docker/packager/packager b/docker/packager/packager index b5bcbada1da..ade36a55591 100755 --- a/docker/packager/packager +++ b/docker/packager/packager @@ -145,6 +145,7 @@ def parse_env_variables( RISCV_SUFFIX = "-riscv64" S390X_SUFFIX = "-s390x" AMD64_COMPAT_SUFFIX = "-amd64-compat" + AMD64_MUSL_SUFFIX = "-amd64-musl" result = [] result.append("OUTPUT_DIR=/output") @@ -163,6 +164,7 @@ def parse_env_variables( is_cross_s390x = compiler.endswith(S390X_SUFFIX) is_cross_freebsd = compiler.endswith(FREEBSD_SUFFIX) is_amd64_compat = compiler.endswith(AMD64_COMPAT_SUFFIX) + is_amd64_musl = compiler.endswith(AMD64_MUSL_SUFFIX) if is_cross_darwin: cc = compiler[: -len(DARWIN_SUFFIX)] @@ -232,6 +234,12 @@ def parse_env_variables( cc = compiler[: -len(AMD64_COMPAT_SUFFIX)] result.append("DEB_ARCH=amd64") cmake_flags.append("-DNO_SSE3_OR_HIGHER=1") + elif is_amd64_musl: + cc = compiler[: -len(AMD64_MUSL_SUFFIX)] + result.append("DEB_ARCH=amd64") + cmake_flags.append( + "-DCMAKE_TOOLCHAIN_FILE=/build/cmake/linux/toolchain-x86_64-musl.cmake" + ) else: cc = compiler result.append("DEB_ARCH=amd64") @@ -396,6 +404,7 @@ def parse_args() -> argparse.Namespace: "clang-17-riscv64", "clang-17-s390x", "clang-17-amd64-compat", + "clang-17-amd64-musl", "clang-17-freebsd", ), default="clang-17", diff --git a/docker/test/fasttest/run.sh b/docker/test/fasttest/run.sh index d3695ba2613..5af05034415 100755 --- a/docker/test/fasttest/run.sh +++ b/docker/test/fasttest/run.sh @@ -16,7 +16,7 @@ export LLVM_VERSION=${LLVM_VERSION:-17} # it being undefined. Also read it as array so that we can pass an empty list # of additional variable to cmake properly, and it doesn't generate an extra # empty parameter. -# Read it as CMAKE_FLAGS to not lose exported FASTTEST_CMAKE_FLAGS on subsequential launch +# Read it as CMAKE_FLAGS to not lose exported FASTTEST_CMAKE_FLAGS on subsequent launch read -ra CMAKE_FLAGS <<< "${FASTTEST_CMAKE_FLAGS:-}" # Run only matching tests. @@ -197,7 +197,7 @@ function run_cmake ( cd "$FASTTEST_BUILD" - cmake "$FASTTEST_SOURCE" -DCMAKE_CXX_COMPILER="clang++-${LLVM_VERSION}" -DCMAKE_C_COMPILER="clang-${LLVM_VERSION}" "${CMAKE_LIBS_CONFIG[@]}" "${CMAKE_FLAGS[@]}" 2>&1 | ts '%Y-%m-%d %H:%M:%S' | tee "$FASTTEST_OUTPUT/cmake_log.txt" + cmake "$FASTTEST_SOURCE" -DCMAKE_CXX_COMPILER="clang++-${LLVM_VERSION}" -DCMAKE_C_COMPILER="clang-${LLVM_VERSION}" -DCMAKE_TOOLCHAIN_FILE="${FASTTEST_SOURCE}/cmake/linux/toolchain-x86_64-musl.cmake" "${CMAKE_LIBS_CONFIG[@]}" "${CMAKE_FLAGS[@]}" 2>&1 | ts '%Y-%m-%d %H:%M:%S' | tee "$FASTTEST_OUTPUT/cmake_log.txt" ) } diff --git a/docs/en/development/build-cross-osx.md b/docs/en/development/build-cross-osx.md index a04d676e92d..eddf24448c1 100644 --- a/docs/en/development/build-cross-osx.md +++ b/docs/en/development/build-cross-osx.md @@ -28,18 +28,20 @@ sudo apt-get install clang-17 Let’s remember the path where we install `cctools` as ${CCTOOLS} ``` bash +mkdir ~/cctools export CCTOOLS=$(cd ~/cctools && pwd) -mkdir ${CCTOOLS} cd ${CCTOOLS} -git clone --depth=1 https://github.com/tpoechtrager/apple-libtapi.git +git clone https://github.com/tpoechtrager/apple-libtapi.git cd apple-libtapi +git checkout 15dfc2a8c9a2a89d06ff227560a69f5265b692f9 INSTALLPREFIX=${CCTOOLS} ./build.sh ./install.sh cd .. -git clone --depth=1 https://github.com/tpoechtrager/cctools-port.git +git clone https://github.com/tpoechtrager/cctools-port.git cd cctools-port/cctools +git checkout 2a3e1c2a6ff54a30f898b70cfb9ba1692a55fad7 ./configure --prefix=$(readlink -f ${CCTOOLS}) --with-libtapi=$(readlink -f ${CCTOOLS}) --target=x86_64-apple-darwin make install ``` diff --git a/docs/en/operations/utilities/clickhouse-local.md b/docs/en/operations/utilities/clickhouse-local.md index c863282efc1..1dac2d25ea5 100644 --- a/docs/en/operations/utilities/clickhouse-local.md +++ b/docs/en/operations/utilities/clickhouse-local.md @@ -216,7 +216,6 @@ Arguments: - `--logger.level` — Log level. - `--ignore-error` — do not stop processing if a query failed. - `-c`, `--config-file` — path to configuration file in same format as for ClickHouse server, by default the configuration empty. -- `--no-system-tables` — do not attach system tables. - `--help` — arguments references for `clickhouse-local`. - `-V`, `--version` — print version information and exit. diff --git a/docs/en/sql-reference/functions/hash-functions.md b/docs/en/sql-reference/functions/hash-functions.md index 912851c9299..a23849c13aa 100644 --- a/docs/en/sql-reference/functions/hash-functions.md +++ b/docs/en/sql-reference/functions/hash-functions.md @@ -319,9 +319,9 @@ This is a relatively fast non-cryptographic hash function of average quality for Calculates a 64-bit hash code from any type of integer. It works faster than intHash32. Average quality. -## SHA1, SHA224, SHA256, SHA512 +## SHA1, SHA224, SHA256, SHA512, SHA512_256 -Calculates SHA-1, SHA-224, SHA-256, SHA-512 hash from a string and returns the resulting set of bytes as [FixedString](/docs/en/sql-reference/data-types/fixedstring.md). +Calculates SHA-1, SHA-224, SHA-256, SHA-512, SHA-512-256 hash from a string and returns the resulting set of bytes as [FixedString](/docs/en/sql-reference/data-types/fixedstring.md). **Syntax** diff --git a/docs/en/sql-reference/statements/create/view.md b/docs/en/sql-reference/statements/create/view.md index 2a8d6788889..56828745048 100644 --- a/docs/en/sql-reference/statements/create/view.md +++ b/docs/en/sql-reference/statements/create/view.md @@ -90,152 +90,11 @@ Views look the same as normal tables. For example, they are listed in the result To delete a view, use [DROP VIEW](../../../sql-reference/statements/drop.md#drop-view). Although `DROP TABLE` works for VIEWs as well. -## Live View [Experimental] +## Live View [Deprecated] -:::note -This is an experimental feature that may change in backwards-incompatible ways in the future releases. Enable usage of live views and `WATCH` query using [allow_experimental_live_view](../../../operations/settings/settings.md#allow-experimental-live-view) setting. Input the command `set allow_experimental_live_view = 1`. -::: +This feature is deprecated and will be removed in the future. -```sql -CREATE LIVE VIEW [IF NOT EXISTS] [db.]table_name [WITH REFRESH [value_in_sec]] AS SELECT ... -``` - -Live views store result of the corresponding [SELECT](../../../sql-reference/statements/select/index.md) query and are updated any time the result of the query changes. Query result as well as partial result needed to combine with new data are stored in memory providing increased performance for repeated queries. Live views can provide push notifications when query result changes using the [WATCH](../../../sql-reference/statements/watch.md) query. - -Live views are triggered by insert into the innermost table specified in the query. - -Live views work similarly to how a query in a distributed table works. But instead of combining partial results from different servers they combine partial result from current data with partial result from the new data. When a live view query includes a subquery then the cached partial result is only stored for the innermost subquery. - -:::info -- [Table function](../../../sql-reference/table-functions/index.md) is not supported as the innermost table. -- Tables that do not have inserts such as a [dictionary](../../../sql-reference/dictionaries/index.md), [system table](../../../operations/system-tables/index.md), a [normal view](#normal), or a [materialized view](#materialized) will not trigger a live view. -- Only queries where one can combine partial result from the old data plus partial result from the new data will work. Live view will not work for queries that require the complete data set to compute the final result or aggregations where the state of the aggregation must be preserved. -- Does not work with replicated or distributed tables where inserts are performed on different nodes. -- Can't be triggered by multiple tables. - -See [WITH REFRESH](#live-view-with-refresh) to force periodic updates of a live view that in some cases can be used as a workaround. -::: - -### Monitoring Live View Changes - -You can monitor changes in the `LIVE VIEW` query result using [WATCH](../../../sql-reference/statements/watch.md) query. - -```sql -WATCH [db.]live_view -``` - -**Example:** - -```sql -CREATE TABLE mt (x Int8) Engine = MergeTree ORDER BY x; -CREATE LIVE VIEW lv AS SELECT sum(x) FROM mt; -``` -Watch a live view while doing a parallel insert into the source table. - -```sql -WATCH lv; -``` - -```bash -┌─sum(x)─┬─_version─┐ -│ 1 │ 1 │ -└────────┴──────────┘ -┌─sum(x)─┬─_version─┐ -│ 3 │ 2 │ -└────────┴──────────┘ -┌─sum(x)─┬─_version─┐ -│ 6 │ 3 │ -└────────┴──────────┘ -``` - -```sql -INSERT INTO mt VALUES (1); -INSERT INTO mt VALUES (2); -INSERT INTO mt VALUES (3); -``` - -Or add [EVENTS](../../../sql-reference/statements/watch.md#events-clause) clause to just get change events. - -```sql -WATCH [db.]live_view EVENTS; -``` - -**Example:** - -```sql -WATCH lv EVENTS; -``` - -```bash -┌─version─┐ -│ 1 │ -└─────────┘ -┌─version─┐ -│ 2 │ -└─────────┘ -┌─version─┐ -│ 3 │ -└─────────┘ -``` - -You can execute [SELECT](../../../sql-reference/statements/select/index.md) query on a live view in the same way as for any regular view or a table. If the query result is cached it will return the result immediately without running the stored query on the underlying tables. - -```sql -SELECT * FROM [db.]live_view WHERE ... -``` - -### Force Live View Refresh - -You can force live view refresh using the `ALTER LIVE VIEW [db.]table_name REFRESH` statement. - -### WITH REFRESH Clause - -When a live view is created with a `WITH REFRESH` clause then it will be automatically refreshed after the specified number of seconds elapse since the last refresh or trigger. - -```sql -CREATE LIVE VIEW [db.]table_name WITH REFRESH [value_in_sec] AS SELECT ... -``` - -If the refresh value is not specified then the value specified by the [periodic_live_view_refresh](../../../operations/settings/settings.md#periodic-live-view-refresh) setting is used. - -**Example:** - -```sql -CREATE LIVE VIEW lv WITH REFRESH 5 AS SELECT now(); -WATCH lv -``` - -```bash -┌───────────────now()─┬─_version─┐ -│ 2021-02-21 08:47:05 │ 1 │ -└─────────────────────┴──────────┘ -┌───────────────now()─┬─_version─┐ -│ 2021-02-21 08:47:10 │ 2 │ -└─────────────────────┴──────────┘ -┌───────────────now()─┬─_version─┐ -│ 2021-02-21 08:47:15 │ 3 │ -└─────────────────────┴──────────┘ -``` - -```sql -WATCH lv -``` - -``` -Code: 60. DB::Exception: Received from localhost:9000. DB::Exception: Table default.lv does not exist.. -``` - -### Live View Usage - -Most common uses of live view tables include: - -- Providing push notifications for query result changes to avoid polling. -- Caching results of most frequent queries to provide immediate query results. -- Watching for table changes and triggering a follow-up select queries. -- Watching metrics from system tables using periodic refresh. - -**See Also** -- [ALTER LIVE VIEW](../alter/view.md#alter-live-view) +For your convenience, the old documentation is located [here](https://pastila.nl/?00f32652/fdf07272a7b54bda7e13b919264e449f.md) ## Window View [Experimental] diff --git a/docs/ru/operations/utilities/clickhouse-local.md b/docs/ru/operations/utilities/clickhouse-local.md index 6f0394a183d..92712a6f6b2 100644 --- a/docs/ru/operations/utilities/clickhouse-local.md +++ b/docs/ru/operations/utilities/clickhouse-local.md @@ -45,7 +45,6 @@ $ clickhouse-local --structure "table_structure" --input-format "format_of_incom - `--logger.level` — уровень логирования. - `--ignore-error` — не прекращать обработку если запрос выдал ошибку. - `-c`, `--config-file` — путь к файлу конфигурации. По умолчанию `clickhouse-local` запускается с пустой конфигурацией. Конфигурационный файл имеет тот же формат, что и для сервера ClickHouse, и в нём можно использовать все конфигурационные параметры сервера. Обычно подключение конфигурации не требуется; если требуется установить отдельный параметр, то это можно сделать ключом с именем параметра. -- `--no-system-tables` — запуск без использования системных таблиц. - `--help` — вывод справочной информации о `clickhouse-local`. - `-V`, `--version` — вывод текущей версии и выход. diff --git a/docs/zh/operations/utilities/clickhouse-local.md b/docs/zh/operations/utilities/clickhouse-local.md index 7428ae06a6e..e8c9503626b 100644 --- a/docs/zh/operations/utilities/clickhouse-local.md +++ b/docs/zh/operations/utilities/clickhouse-local.md @@ -45,7 +45,6 @@ clickhouse-local --structure "table_structure" --input-format "format_of_incomin - `--logger.level` — 日志级别。 - `--ignore-error` — 当查询失败时，不停止处理。 - `-c`, `--config-file` — 与ClickHouse服务器格式相同配置文件的路径，默认情况下配置为空。 -- `--no-system-tables` — 不附加系统表。 - `--help` — `clickhouse-local`使用帮助信息。 - `-V`, `--version` — 打印版本信息并退出。 diff --git a/programs/keeper/Keeper.cpp b/programs/keeper/Keeper.cpp index e04e669abae..1acf7e39b04 100644 --- a/programs/keeper/Keeper.cpp +++ b/programs/keeper/Keeper.cpp @@ -328,6 +328,13 @@ try config().getUInt("max_thread_pool_free_size", 1000), config().getUInt("thread_pool_queue_size", 10000) ); + /// Wait for all threads to avoid possible use-after-free (for example logging objects can be already destroyed). + SCOPE_EXIT({ + Stopwatch watch; + LOG_INFO(log, "Waiting for background threads"); + GlobalThreadPool::instance().shutdown(); + LOG_INFO(log, "Background threads finished in {} ms", watch.elapsedMilliseconds()); + }); static ServerErrorHandler error_handler; Poco::ErrorHandler::set(&error_handler); diff --git a/programs/local/LocalServer.cpp b/programs/local/LocalServer.cpp index fbb64ea1135..8e526812957 100644 --- a/programs/local/LocalServer.cpp +++ b/programs/local/LocalServer.cpp @@ -744,7 +744,7 @@ void LocalServer::processConfig() LOG_DEBUG(log, "Loading metadata from {}", path); auto startup_system_tasks = loadMetadataSystem(global_context); - attachSystemTablesLocal(global_context, *createMemoryDatabaseIfNotExists(global_context, DatabaseCatalog::SYSTEM_DATABASE)); + attachSystemTablesLocal(global_context, *createMemoryDatabaseIfNotExists(global_context, DatabaseCatalog::SYSTEM_DATABASE)); attachInformationSchema(global_context, *createMemoryDatabaseIfNotExists(global_context, DatabaseCatalog::INFORMATION_SCHEMA)); attachInformationSchema(global_context, *createMemoryDatabaseIfNotExists(global_context, DatabaseCatalog::INFORMATION_SCHEMA_UPPERCASE)); waitLoad(TablesLoaderForegroundPoolId, startup_system_tasks); @@ -761,9 +761,9 @@ void LocalServer::processConfig() LOG_DEBUG(log, "Loaded metadata."); } - else if (!config().has("no-system-tables")) + else { - attachSystemTablesLocal(global_context, *createMemoryDatabaseIfNotExists(global_context, DatabaseCatalog::SYSTEM_DATABASE)); + attachSystemTablesLocal(global_context, *createMemoryDatabaseIfNotExists(global_context, DatabaseCatalog::SYSTEM_DATABASE)); attachInformationSchema(global_context, *createMemoryDatabaseIfNotExists(global_context, DatabaseCatalog::INFORMATION_SCHEMA)); attachInformationSchema(global_context, *createMemoryDatabaseIfNotExists(global_context, DatabaseCatalog::INFORMATION_SCHEMA_UPPERCASE)); } @@ -842,7 +842,6 @@ void LocalServer::addOptions(OptionsDescription & options_description) ("logger.log", po::value(), "Log file name") ("logger.level", po::value(), "Log level") - ("no-system-tables", "do not attach system tables (better startup time)") ("path", po::value(), "Storage path") ("only-system-tables", "attach only system tables from specified path") ("top_level_domains_path", po::value(), "Path to lists with custom TLDs") @@ -871,8 +870,6 @@ void LocalServer::processOptions(const OptionsDescription &, const CommandLineOp config().setString("table-file", options["file"].as()); if (options.count("structure")) config().setString("table-structure", options["structure"].as()); - if (options.count("no-system-tables")) - config().setBool("no-system-tables", true); if (options.count("only-system-tables")) config().setBool("only-system-tables", true); if (options.count("database")) diff --git a/programs/server/Server.cpp b/programs/server/Server.cpp index de0cfb9b9fa..8076d108083 100644 --- a/programs/server/Server.cpp +++ b/programs/server/Server.cpp @@ -657,6 +657,11 @@ try CurrentMetrics::set(CurrentMetrics::Revision, ClickHouseRevision::getVersionRevision()); CurrentMetrics::set(CurrentMetrics::VersionInteger, ClickHouseRevision::getVersionInteger()); + Poco::ThreadPool server_pool(3, server_settings.max_connections); + std::mutex servers_lock; + std::vector servers; + std::vector servers_to_start_before_tables; + /** Context contains all that query execution is dependent: * settings, available functions, data types, aggregate functions, databases, ... */ @@ -697,6 +702,68 @@ try server_settings.max_thread_pool_size, server_settings.max_thread_pool_free_size, server_settings.thread_pool_queue_size); + /// Wait for all threads to avoid possible use-after-free (for example logging objects can be already destroyed). + SCOPE_EXIT({ + Stopwatch watch; + LOG_INFO(log, "Waiting for background threads"); + GlobalThreadPool::instance().shutdown(); + LOG_INFO(log, "Background threads finished in {} ms", watch.elapsedMilliseconds()); + }); + + /// NOTE: global context should be destroyed *before* GlobalThreadPool::shutdown() + /// Otherwise GlobalThreadPool::shutdown() will hang, since Context holds some threads. + SCOPE_EXIT({ + /** Ask to cancel background jobs all table engines, + * and also query_log. + * It is important to do early, not in destructor of Context, because + * table engines could use Context on destroy. + */ + LOG_INFO(log, "Shutting down storages."); + + global_context->shutdown(); + + LOG_DEBUG(log, "Shut down storages."); + + if (!servers_to_start_before_tables.empty()) + { + LOG_DEBUG(log, "Waiting for current connections to servers for tables to finish."); + size_t current_connections = 0; + { + std::lock_guard lock(servers_lock); + for (auto & server : servers_to_start_before_tables) + { + server.stop(); + current_connections += server.currentConnections(); + } + } + + if (current_connections) + LOG_INFO(log, "Closed all listening sockets. Waiting for {} outstanding connections.", current_connections); + else + LOG_INFO(log, "Closed all listening sockets."); + + if (current_connections > 0) + current_connections = waitServersToFinish(servers_to_start_before_tables, servers_lock, server_settings.shutdown_wait_unfinished); + + if (current_connections) + LOG_INFO(log, "Closed connections to servers for tables. But {} remain. Probably some tables of other users cannot finish their connections after context shutdown.", current_connections); + else + LOG_INFO(log, "Closed connections to servers for tables."); + } + + global_context->shutdownKeeperDispatcher(); + + /// Wait server pool to avoid use-after-free of destroyed context in the handlers + server_pool.joinAll(); + + /** Explicitly destroy Context. It is more convenient than in destructor of Server, because logger is still available. + * At this moment, no one could own shared part of Context. + */ + global_context.reset(); + shared_context.reset(); + LOG_DEBUG(log, "Destroyed global context."); + }); + #if USE_AZURE_BLOB_STORAGE /// It makes sense to deinitialize libxml after joining of all threads @@ -755,10 +822,6 @@ try } } - Poco::ThreadPool server_pool(3, server_settings.max_connections); - std::mutex servers_lock; - std::vector servers; - std::vector servers_to_start_before_tables; /// This object will periodically calculate some metrics. ServerAsynchronousMetrics async_metrics( global_context, @@ -1598,60 +1661,6 @@ try /// try set up encryption. There are some errors in config, error will be printed and server wouldn't start. CompressionCodecEncrypted::Configuration::instance().load(config(), "encryption_codecs"); - SCOPE_EXIT({ - async_metrics.stop(); - - /** Ask to cancel background jobs all table engines, - * and also query_log. - * It is important to do early, not in destructor of Context, because - * table engines could use Context on destroy. - */ - LOG_INFO(log, "Shutting down storages."); - - global_context->shutdown(); - - LOG_DEBUG(log, "Shut down storages."); - - if (!servers_to_start_before_tables.empty()) - { - LOG_DEBUG(log, "Waiting for current connections to servers for tables to finish."); - size_t current_connections = 0; - { - std::lock_guard lock(servers_lock); - for (auto & server : servers_to_start_before_tables) - { - server.stop(); - current_connections += server.currentConnections(); - } - } - - if (current_connections) - LOG_INFO(log, "Closed all listening sockets. Waiting for {} outstanding connections.", current_connections); - else - LOG_INFO(log, "Closed all listening sockets."); - - if (current_connections > 0) - current_connections = waitServersToFinish(servers_to_start_before_tables, servers_lock, server_settings.shutdown_wait_unfinished); - - if (current_connections) - LOG_INFO(log, "Closed connections to servers for tables. But {} remain. Probably some tables of other users cannot finish their connections after context shutdown.", current_connections); - else - LOG_INFO(log, "Closed connections to servers for tables."); - - global_context->shutdownKeeperDispatcher(); - } - - /// Wait server pool to avoid use-after-free of destroyed context in the handlers - server_pool.joinAll(); - - /** Explicitly destroy Context. It is more convenient than in destructor of Server, because logger is still available. - * At this moment, no one could own shared part of Context. - */ - global_context.reset(); - shared_context.reset(); - LOG_DEBUG(log, "Destroyed global context."); - }); - /// DNSCacheUpdater uses BackgroundSchedulePool which lives in shared context /// and thus this object must be created after the SCOPE_EXIT object where shared /// context is destroyed. diff --git a/src/AggregateFunctions/AggregateFunctionAny.cpp b/src/AggregateFunctions/AggregateFunctionAny.cpp index 15681eca817..a6010ff07c3 100644 --- a/src/AggregateFunctions/AggregateFunctionAny.cpp +++ b/src/AggregateFunctions/AggregateFunctionAny.cpp @@ -110,7 +110,7 @@ public: } } } - else + else if (row_begin < row_end) { size_t pos = First ? row_begin : row_end - 1; add(place, columns, pos, arena); diff --git a/src/AggregateFunctions/IAggregateFunction.h b/src/AggregateFunctions/IAggregateFunction.h index 2a62b76adbb..a8254baac3a 100644 --- a/src/AggregateFunctions/IAggregateFunction.h +++ b/src/AggregateFunctions/IAggregateFunction.h @@ -549,8 +549,10 @@ public: auto to = std::lower_bound(offsets.begin(), offsets.end(), row_end) - offsets.begin() + 1; size_t num_defaults = (row_end - row_begin) - (to - from); - static_cast(this)->addBatchSinglePlace(from, to, place, &values, arena, -1); - static_cast(this)->addManyDefaults(place, &values, num_defaults, arena); + if (from < to) + static_cast(this)->addBatchSinglePlace(from, to, place, &values, arena, -1); + if (num_defaults > 0) + static_cast(this)->addManyDefaults(place, &values, num_defaults, arena); } void addBatchSinglePlaceNotNull( /// NOLINT diff --git a/src/Client/LineReader.cpp b/src/Client/LineReader.cpp index 2ec90240fd1..b3559657ced 100644 --- a/src/Client/LineReader.cpp +++ b/src/Client/LineReader.cpp @@ -7,7 +7,7 @@ #include #include #include -#include +#include #include #include @@ -27,11 +27,8 @@ void trim(String & s) /// Allows delaying the start of query execution until the entirety of query is inserted. bool hasInputData() { - timeval timeout = {0, 0}; - fd_set fds{}; - FD_ZERO(&fds); - FD_SET(STDIN_FILENO, &fds); - return select(1, &fds, nullptr, nullptr, &timeout) == 1; + pollfd fd{STDIN_FILENO, POLLIN, 0}; + return poll(&fd, 1, 0) == 1; } struct NoCaseCompare diff --git a/src/Common/ProfileEvents.cpp b/src/Common/ProfileEvents.cpp index 9908b037238..1054eae4088 100644 --- a/src/Common/ProfileEvents.cpp +++ b/src/Common/ProfileEvents.cpp @@ -238,6 +238,7 @@ M(DictCacheLockReadNs, "Number of nanoseconds spend in waiting for read lock to lookup the data for the dictionaries of 'cache' types.") \ \ M(DistributedSyncInsertionTimeoutExceeded, "A timeout has exceeded while waiting for shards during synchronous insertion into a Distributed table (with 'distributed_foreground_insert' = 1)") \ + M(DistributedAsyncInsertionFailures, "Number of failures for asynchronous insertion into a Distributed table (with 'distributed_foreground_insert' = 0)") \ M(DataAfterMergeDiffersFromReplica, R"( Number of times data after merge is not byte-identical to the data on another replicas. There could be several reasons: 1. Using newer version of compression library after server update. diff --git a/src/Common/ThreadPool.cpp b/src/Common/ThreadPool.cpp index 8cba13373b9..565affb0c65 100644 --- a/src/Common/ThreadPool.cpp +++ b/src/Common/ThreadPool.cpp @@ -500,3 +500,10 @@ GlobalThreadPool & GlobalThreadPool::instance() return *the_instance; } +void GlobalThreadPool::shutdown() +{ + if (the_instance) + { + the_instance->finalize(); + } +} diff --git a/src/Common/ThreadPool.h b/src/Common/ThreadPool.h index c8eefedd838..3117509ab8f 100644 --- a/src/Common/ThreadPool.h +++ b/src/Common/ThreadPool.h @@ -109,6 +109,8 @@ public: void addOnDestroyCallback(OnDestroyCallback && callback); private: + friend class GlobalThreadPool; + mutable std::mutex mutex; std::condition_variable job_finished; std::condition_variable new_job_or_shutdown; @@ -205,6 +207,7 @@ class GlobalThreadPool : public FreeThreadPool, private boost::noncopyable public: static void initialize(size_t max_threads = 10000, size_t max_free_threads = 1000, size_t queue_size = 10000); static GlobalThreadPool & instance(); + static void shutdown(); }; diff --git a/src/Common/mysqlxx/Pool.cpp b/src/Common/mysqlxx/Pool.cpp index 43df0aa6708..cc5b18214c8 100644 --- a/src/Common/mysqlxx/Pool.cpp +++ b/src/Common/mysqlxx/Pool.cpp @@ -52,8 +52,7 @@ void Pool::Entry::decrementRefCount() Pool::Pool(const Poco::Util::AbstractConfiguration & cfg, const std::string & config_name, unsigned default_connections_, unsigned max_connections_, const char * parent_config_name_) - : logger(Poco::Logger::get("mysqlxx::Pool")) - , default_connections(default_connections_) + : default_connections(default_connections_) , max_connections(max_connections_) { server = cfg.getString(config_name + ".host"); @@ -127,6 +126,38 @@ Pool::Pool(const Poco::Util::AbstractConfiguration & cfg, const std::string & co } +Pool::Pool( + const std::string & db_, + const std::string & server_, + const std::string & user_, + const std::string & password_, + unsigned port_, + const std::string & socket_, + unsigned connect_timeout_, + unsigned rw_timeout_, + unsigned default_connections_, + unsigned max_connections_, + unsigned enable_local_infile_, + bool opt_reconnect_) + : default_connections(default_connections_) + , max_connections(max_connections_) + , db(db_) + , server(server_) + , user(user_) + , password(password_) + , port(port_) + , socket(socket_) + , connect_timeout(connect_timeout_) + , rw_timeout(rw_timeout_) + , enable_local_infile(enable_local_infile_) + , opt_reconnect(opt_reconnect_) +{ + LOG_DEBUG(log, + "Created MySQL Pool with settings: connect_timeout={}, read_write_timeout={}, default_connections_number={}, max_connections_number={}", + connect_timeout, rw_timeout, default_connections, max_connections); +} + + Pool::~Pool() { std::lock_guard lock(mutex); @@ -148,29 +179,29 @@ Pool::Entry Pool::get(uint64_t wait_timeout) initialize(); for (;;) { - logger.trace("(%s): Iterating through existing MySQL connections", getDescription()); + LOG_TRACE(log, "{}: Iterating through existing MySQL connections", getDescription()); for (auto & connection : connections) { if (connection->ref_count == 0) { - logger.test("Found free connection in pool, returning it to the caller"); + LOG_TEST(log, "Found free connection in pool, returning it to the caller"); return Entry(connection, this); } } - logger.trace("(%s): Trying to allocate a new connection.", getDescription()); + LOG_TRACE(log, "{}: Trying to allocate a new connection.", getDescription()); if (connections.size() < static_cast(max_connections)) { Connection * conn = allocConnection(); if (conn) return Entry(conn, this); - logger.trace("(%s): Unable to create a new connection: Allocation failed.", getDescription()); + LOG_TRACE(log, "{}: Unable to create a new connection: Allocation failed.", getDescription()); } else { - logger.trace("(%s): Unable to create a new connection: Max number of connections has been reached.", getDescription()); + LOG_TRACE(log, "{}: Unable to create a new connection: Max number of connections has been reached.", getDescription()); } if (!wait_timeout) @@ -180,7 +211,7 @@ Pool::Entry Pool::get(uint64_t wait_timeout) throw Poco::Exception("mysqlxx::Pool is full (connection_wait_timeout is exceeded)"); lock.unlock(); - logger.trace("(%s): Sleeping for %d seconds.", getDescription(), MYSQLXX_POOL_SLEEP_ON_CONNECT_FAIL); + LOG_TRACE(log, "{}: Sleeping for {} seconds.", getDescription(), MYSQLXX_POOL_SLEEP_ON_CONNECT_FAIL); sleepForSeconds(MYSQLXX_POOL_SLEEP_ON_CONNECT_FAIL); lock.lock(); } @@ -206,7 +237,7 @@ Pool::Entry Pool::tryGet() return res; } - logger.debug("(%s): Idle connection to MySQL server cannot be recovered, dropping it.", getDescription()); + LOG_DEBUG(log, "{}: Idle connection to MySQL server cannot be recovered, dropping it.", getDescription()); /// This one is disconnected, cannot be reestablished and so needs to be disposed of. connection_it = connections.erase(connection_it); @@ -229,7 +260,7 @@ Pool::Entry Pool::tryGet() void Pool::removeConnection(Connection* connection) { - logger.trace("(%s): Removing connection.", getDescription()); + LOG_TRACE(log, "{}: Removing connection.", getDescription()); std::lock_guard lock(mutex); if (connection) @@ -260,8 +291,8 @@ void Pool::Entry::forceConnected() const else sleepForSeconds(MYSQLXX_POOL_SLEEP_ON_CONNECT_FAIL); - pool->logger.debug( - "Creating a new MySQL connection to %s with settings: connect_timeout=%u, read_write_timeout=%u", + LOG_DEBUG(pool->log, + "Creating a new MySQL connection to {} with settings: connect_timeout={}, read_write_timeout={}", pool->description, pool->connect_timeout, pool->rw_timeout); data->conn.connect( @@ -287,21 +318,21 @@ bool Pool::Entry::tryForceConnected() const auto * const mysql_driver = data->conn.getDriver(); const auto prev_connection_id = mysql_thread_id(mysql_driver); - pool->logger.trace("Entry(connection %lu): sending PING to check if it is alive.", prev_connection_id); + LOG_TRACE(pool->log, "Entry(connection {}): sending PING to check if it is alive.", prev_connection_id); if (data->conn.ping()) /// Attempts to reestablish lost connection { const auto current_connection_id = mysql_thread_id(mysql_driver); if (prev_connection_id != current_connection_id) { - pool->logger.debug("Entry(connection %lu): Reconnected to MySQL server. Connection id changed: %lu -> %lu", - current_connection_id, prev_connection_id, current_connection_id); + LOG_DEBUG(pool->log, "Entry(connection {}): Reconnected to MySQL server. Connection id changed: {} -> {}", + current_connection_id, prev_connection_id, current_connection_id); } - pool->logger.trace("Entry(connection %lu): PING ok.", current_connection_id); + LOG_TRACE(pool->log, "Entry(connection {}): PING ok.", current_connection_id); return true; } - pool->logger.trace("Entry(connection %lu): PING failed.", prev_connection_id); + LOG_TRACE(pool->log, "Entry(connection {}): PING failed.", prev_connection_id); return false; } @@ -326,10 +357,10 @@ Pool::Connection * Pool::allocConnection(bool dont_throw_if_failed_first_time) try { - logger.debug("Connecting to %s", description); + LOG_DEBUG(log, "Connecting to {}", description); - logger.debug( - "Creating a new MySQL connection to %s with settings: connect_timeout=%u, read_write_timeout=%u", + LOG_DEBUG(log, + "Creating a new MySQL connection to {} with settings: connect_timeout={}, read_write_timeout={}", description, connect_timeout, rw_timeout); conn_ptr->conn.connect( @@ -349,7 +380,7 @@ Pool::Connection * Pool::allocConnection(bool dont_throw_if_failed_first_time) } catch (mysqlxx::ConnectionFailed & e) { - logger.error(e.what()); + LOG_ERROR(log, "Failed to connect to MySQL ({}): {}", description, e.what()); if ((!was_successful && !dont_throw_if_failed_first_time) || e.errnum() == ER_ACCESS_DENIED_ERROR diff --git a/src/Common/mysqlxx/mysqlxx/Pool.h b/src/Common/mysqlxx/mysqlxx/Pool.h index 52d116e39ce..bb4d0cefbdc 100644 --- a/src/Common/mysqlxx/mysqlxx/Pool.h +++ b/src/Common/mysqlxx/mysqlxx/Pool.h @@ -169,28 +169,10 @@ public: unsigned default_connections_ = MYSQLXX_POOL_DEFAULT_START_CONNECTIONS, unsigned max_connections_ = MYSQLXX_POOL_DEFAULT_MAX_CONNECTIONS, unsigned enable_local_infile_ = MYSQLXX_DEFAULT_ENABLE_LOCAL_INFILE, - bool opt_reconnect_ = MYSQLXX_DEFAULT_MYSQL_OPT_RECONNECT) - : logger(Poco::Logger::get("mysqlxx::Pool")) - , default_connections(default_connections_) - , max_connections(max_connections_) - , db(db_) - , server(server_) - , user(user_) - , password(password_) - , port(port_) - , socket(socket_) - , connect_timeout(connect_timeout_) - , rw_timeout(rw_timeout_) - , enable_local_infile(enable_local_infile_) - , opt_reconnect(opt_reconnect_) - { - logger.debug( - "Created MySQL Pool with settings: connect_timeout=%u, read_write_timeout=%u, default_connections_number=%u, max_connections_number=%u", - connect_timeout, rw_timeout, default_connections, max_connections); - } + bool opt_reconnect_ = MYSQLXX_DEFAULT_MYSQL_OPT_RECONNECT); Pool(const Pool & other) - : logger(other.logger), default_connections{other.default_connections}, + : default_connections{other.default_connections}, max_connections{other.max_connections}, db{other.db}, server{other.server}, user{other.user}, password{other.password}, @@ -220,7 +202,7 @@ public: void removeConnection(Connection * connection); protected: - Poco::Logger & logger; + Poco::Logger * log = &Poco::Logger::get("mysqlxx::Pool"); /// Number of MySQL connections which are created at launch. unsigned default_connections; diff --git a/src/Coordination/CoordinationSettings.h b/src/Coordination/CoordinationSettings.h index a58f2b04797..13ef304b353 100644 --- a/src/Coordination/CoordinationSettings.h +++ b/src/Coordination/CoordinationSettings.h @@ -43,6 +43,7 @@ struct Settings; M(UInt64, max_requests_batch_bytes_size, 100*1024, "Max size in bytes of batch of requests that can be sent to RAFT", 0) \ M(UInt64, max_flush_batch_size, 1000, "Max size of batch of requests that can be flushed together", 0) \ M(UInt64, max_requests_quick_batch_size, 100, "Max size of batch of requests to try to get before proceeding with RAFT. Keeper will not wait for requests but take only requests that are already in queue" , 0) \ + M(UInt64, max_memory_usage_soft_limit, 0, "Soft limit in bytes of keeper memory usage", 0) \ M(Bool, quorum_reads, false, "Execute read requests as writes through whole RAFT consesus with similar speed", 0) \ M(Bool, force_sync, true, "Call fsync on each change in RAFT changelog", 0) \ M(Bool, compress_logs, false, "Write compressed coordination logs in ZSTD format", 0) \ diff --git a/src/Coordination/KeeperAsynchronousMetrics.cpp b/src/Coordination/KeeperAsynchronousMetrics.cpp index 890079e98f7..4471012e917 100644 --- a/src/Coordination/KeeperAsynchronousMetrics.cpp +++ b/src/Coordination/KeeperAsynchronousMetrics.cpp @@ -113,6 +113,12 @@ KeeperAsynchronousMetrics::KeeperAsynchronousMetrics( { } +KeeperAsynchronousMetrics::~KeeperAsynchronousMetrics() +{ + /// NOTE: stop() from base class is not enough, since this leads to leak on vptr + stop(); +} + void KeeperAsynchronousMetrics::updateImpl(AsynchronousMetricValues & new_values, TimePoint /*update_time*/, TimePoint /*current_time*/) { #if USE_NURAFT diff --git a/src/Coordination/KeeperAsynchronousMetrics.h b/src/Coordination/KeeperAsynchronousMetrics.h index 14092c11c15..457a7112507 100644 --- a/src/Coordination/KeeperAsynchronousMetrics.h +++ b/src/Coordination/KeeperAsynchronousMetrics.h @@ -14,6 +14,7 @@ class KeeperAsynchronousMetrics : public AsynchronousMetrics public: KeeperAsynchronousMetrics( ContextPtr context_, int update_period_seconds, const ProtocolServerMetricsFunc & protocol_server_metrics_func_); + ~KeeperAsynchronousMetrics() override; private: ContextPtr context; diff --git a/src/Coordination/KeeperDispatcher.cpp b/src/Coordination/KeeperDispatcher.cpp index f69a9c11e97..1299e9c9f20 100644 --- a/src/Coordination/KeeperDispatcher.cpp +++ b/src/Coordination/KeeperDispatcher.cpp @@ -51,6 +51,56 @@ namespace ErrorCodes extern const int SYSTEM_ERROR; } +namespace +{ + +bool checkIfRequestIncreaseMem(const Coordination::ZooKeeperRequestPtr & request) +{ + if (request->getOpNum() == Coordination::OpNum::Create + || request->getOpNum() == Coordination::OpNum::CreateIfNotExists + || request->getOpNum() == Coordination::OpNum::Set) + { + return true; + } + else if (request->getOpNum() == Coordination::OpNum::Multi) + { + Coordination::ZooKeeperMultiRequest & multi_req = dynamic_cast(*request); + Int64 memory_delta = 0; + for (const auto & sub_req : multi_req.requests) + { + auto sub_zk_request = std::dynamic_pointer_cast(sub_req); + switch (sub_zk_request->getOpNum()) + { + case Coordination::OpNum::Create: + case Coordination::OpNum::CreateIfNotExists: + { + Coordination::ZooKeeperCreateRequest & create_req = dynamic_cast(*sub_zk_request); + memory_delta += create_req.bytesSize(); + break; + } + case Coordination::OpNum::Set: + { + Coordination::ZooKeeperSetRequest & set_req = dynamic_cast(*sub_zk_request); + memory_delta += set_req.bytesSize(); + break; + } + case Coordination::OpNum::Remove: + { + Coordination::ZooKeeperRemoveRequest & remove_req = dynamic_cast(*sub_zk_request); + memory_delta -= remove_req.bytesSize(); + break; + } + default: + break; + } + } + return memory_delta > 0; + } + + return false; +} + +} KeeperDispatcher::KeeperDispatcher() : responses_queue(std::numeric_limits::max()) @@ -93,6 +143,14 @@ void KeeperDispatcher::requestThread() if (shutdown_called) break; + Int64 mem_soft_limit = configuration_and_settings->coordination_settings->max_memory_usage_soft_limit; + if (configuration_and_settings->standalone_keeper && mem_soft_limit > 0 && total_memory_tracker.get() >= mem_soft_limit && checkIfRequestIncreaseMem(request.request)) + { + LOG_TRACE(log, "Processing requests refused because of max_memory_usage_soft_limit {}, the total used memory is {}, request type is {}", mem_soft_limit, total_memory_tracker.get(), request.request->getOpNum()); + addErrorResponses({request}, Coordination::Error::ZCONNECTIONLOSS); + continue; + } + KeeperStorage::RequestsForSessions current_batch; size_t current_batch_bytes_size = 0; diff --git a/src/Coordination/Standalone/Context.cpp b/src/Coordination/Standalone/Context.cpp index dba4a8934b9..97a034b22a4 100644 --- a/src/Coordination/Standalone/Context.cpp +++ b/src/Coordination/Standalone/Context.cpp @@ -4,6 +4,7 @@ #include #include #include +#include #include @@ -62,6 +63,11 @@ struct ContextSharedPart : boost::noncopyable mutable std::unique_ptr asynchronous_local_fs_reader; mutable std::unique_ptr synchronous_local_fs_reader; +#if USE_LIBURING + mutable OnceFlag io_uring_reader_initialized; + mutable std::unique_ptr io_uring_reader; +#endif + mutable OnceFlag threadpool_writer_initialized; mutable std::unique_ptr threadpool_writer; @@ -225,6 +231,17 @@ IAsynchronousReader & Context::getThreadPoolReader(FilesystemReaderType type) co } } +#if USE_LIBURING +IOUringReader & Context::getIOURingReader() const +{ + callOnce(shared->io_uring_reader_initialized, [&] { + shared->io_uring_reader = std::make_unique(512); + }); + + return *shared->io_uring_reader; +} +#endif + std::shared_ptr Context::getFilesystemCacheLog() const { return nullptr; diff --git a/src/Coordination/Standalone/Context.h b/src/Coordination/Standalone/Context.h index 7555618233d..5aecf850d7c 100644 --- a/src/Coordination/Standalone/Context.h +++ b/src/Coordination/Standalone/Context.h @@ -20,6 +20,8 @@ #include +#include "config.h" + namespace DB { @@ -28,6 +30,7 @@ class Macros; class FilesystemCacheLog; class FilesystemReadPrefetchesLog; class BlobStorageLog; +class IOUringReader; /// A small class which owns ContextShared. /// We don't use something like unique_ptr directly to allow ContextShared type to be incomplete. @@ -127,6 +130,9 @@ public: ApplicationType getApplicationType() const { return ApplicationType::KEEPER; } IAsynchronousReader & getThreadPoolReader(FilesystemReaderType type) const; +#if USE_LIBURING + IOUringReader & getIOURingReader() const; +#endif std::shared_ptr getAsyncReadCounters() const; ThreadPool & getThreadPoolWriter() const; diff --git a/src/Core/Settings.h b/src/Core/Settings.h index a6280a28436..9601cd3e398 100644 --- a/src/Core/Settings.h +++ b/src/Core/Settings.h @@ -156,6 +156,7 @@ class IColumn; M(Bool, allow_suspicious_low_cardinality_types, false, "In CREATE TABLE statement allows specifying LowCardinality modifier for types of small fixed size (8 or less). Enabling this may increase merge times and memory consumption.", 0) \ M(Bool, allow_suspicious_fixed_string_types, false, "In CREATE TABLE statement allows creating columns of type FixedString(n) with n > 256. FixedString with length >= 256 is suspicious and most likely indicates misusage", 0) \ M(Bool, allow_suspicious_indices, false, "Reject primary/secondary indexes and sorting keys with identical expressions", 0) \ + M(Bool, allow_suspicious_ttl_expressions, false, "Reject TTL expressions that don't depend on any of table's columns. It indicates a user error most of the time.", 0) \ M(Bool, compile_expressions, false, "Compile some scalar functions and operators to native code.", 0) \ M(UInt64, min_count_to_compile_expression, 3, "The number of identical expressions before they are JIT-compiled", 0) \ M(Bool, compile_aggregate_expressions, true, "Compile aggregate functions to native code.", 0) \ diff --git a/src/Core/SettingsChangesHistory.h b/src/Core/SettingsChangesHistory.h index 90a18b9b2f5..54c28fb9f92 100644 --- a/src/Core/SettingsChangesHistory.h +++ b/src/Core/SettingsChangesHistory.h @@ -7,6 +7,7 @@ #include #include + namespace DB { @@ -80,6 +81,7 @@ namespace SettingsChangesHistory /// It's used to implement `compatibility` setting (see https://github.com/ClickHouse/ClickHouse/issues/35972) static std::map settings_changes_history = { + {"23.12", {{"allow_suspicious_ttl_expressions", true, false, "It is a new setting, and in previous versions the behavior was equivalent to allowing."}}}, {"23.9", {{"optimize_group_by_constant_keys", false, true, "Optimize group by constant keys by default"}, {"input_format_json_try_infer_named_tuples_from_objects", false, true, "Try to infer named Tuples from JSON objects by default"}, {"input_format_json_read_numbers_as_strings", false, true, "Allow to read numbers as strings in JSON formats by default"}, diff --git a/src/Databases/DatabaseAtomic.cpp b/src/Databases/DatabaseAtomic.cpp index 1daa6351c23..1d78e4aff67 100644 --- a/src/Databases/DatabaseAtomic.cpp +++ b/src/Databases/DatabaseAtomic.cpp @@ -89,15 +89,14 @@ void DatabaseAtomic::drop(ContextPtr) fs::remove_all(getMetadataPath()); } -void DatabaseAtomic::attachTable(ContextPtr /* context_ */, const String & name, const StoragePtr & table, const String & relative_table_path) +void DatabaseAtomic::attachTableUnlocked(ContextPtr local_context, const String & name, const StoragePtr & table, const String & relative_table_path) { assert(relative_table_path != data_path && !relative_table_path.empty()); DetachedTables not_in_use; - std::lock_guard lock(mutex); not_in_use = cleanupDetachedTables(); auto table_id = table->getStorageID(); assertDetachedTableNotInUse(table_id.uuid); - DatabaseOrdinary::attachTableUnlocked(name, table); + DatabaseOrdinary::attachTableUnlocked(local_context, name, table, relative_table_path); table_name_to_path.emplace(std::make_pair(name, relative_table_path)); } @@ -325,7 +324,7 @@ void DatabaseAtomic::commitCreateTable(const ASTCreateQuery & query, const Stora /// It throws if `table_metadata_path` already exists (it's possible if table was detached) renameNoReplace(table_metadata_tmp_path, table_metadata_path); /// Commit point (a sort of) - attachTableUnlocked(query.getTable(), table); /// Should never throw + DatabaseWithOwnTablesBase::attachTableUnlocked(query_context, query.getTable(), table, /*relative_table_path=*/ {}); /// Should never throw table_name_to_path.emplace(query.getTable(), table_data_path); } catch (...) diff --git a/src/Databases/DatabaseAtomic.h b/src/Databases/DatabaseAtomic.h index 83cb51be1ff..1eb6dd1cbbd 100644 --- a/src/Databases/DatabaseAtomic.h +++ b/src/Databases/DatabaseAtomic.h @@ -38,7 +38,6 @@ public: void dropTable(ContextPtr context, const String & table_name, bool sync) override; void dropTableImpl(ContextPtr context, const String & table_name, bool sync); - void attachTable(ContextPtr context, const String & name, const StoragePtr & table, const String & relative_table_path) override; StoragePtr detachTable(ContextPtr context, const String & name) override; String getTableDataPath(const String & table_name) const override; @@ -66,6 +65,8 @@ public: void setDetachedTableNotInUseForce(const UUID & uuid) override; protected: + void attachTableUnlocked(ContextPtr local_context, const String & name, const StoragePtr & table, const String & relative_table_path) TSA_REQUIRES(mutex) override; + void commitAlterTable(const StorageID & table_id, const String & table_metadata_tmp_path, const String & table_metadata_path, const String & statement, ContextPtr query_context) override; void commitCreateTable(const ASTCreateQuery & query, const StoragePtr & table, const String & table_metadata_tmp_path, const String & table_metadata_path, ContextPtr query_context) override; diff --git a/src/Databases/DatabaseLazy.cpp b/src/Databases/DatabaseLazy.cpp index 896ae99656f..ec9e1e5ee57 100644 --- a/src/Databases/DatabaseLazy.cpp +++ b/src/Databases/DatabaseLazy.cpp @@ -161,10 +161,9 @@ bool DatabaseLazy::empty() const return tables_cache.empty(); } -void DatabaseLazy::attachTable(ContextPtr /* context_ */, const String & table_name, const StoragePtr & table, const String &) +void DatabaseLazy::attachTableUnlocked(ContextPtr /* context_ */, const String & table_name, const StoragePtr & table, const String &) { LOG_DEBUG(log, "Attach table {}.", backQuote(table_name)); - std::lock_guard lock(mutex); time_t current_time = std::chrono::system_clock::to_time_t(std::chrono::system_clock::now()); auto [it, inserted] = tables_cache.emplace(std::piecewise_construct, diff --git a/src/Databases/DatabaseLazy.h b/src/Databases/DatabaseLazy.h index 2b1b119754d..370c49557de 100644 --- a/src/Databases/DatabaseLazy.h +++ b/src/Databases/DatabaseLazy.h @@ -64,14 +64,15 @@ public: DatabaseTablesIteratorPtr getTablesIterator(ContextPtr context, const FilterByNameFunction & filter_by_table_name) const override; - void attachTable(ContextPtr context, const String & table_name, const StoragePtr & table, const String & relative_table_path) override; - StoragePtr detachTable(ContextPtr context, const String & table_name) override; void shutdown() override; ~DatabaseLazy() override; +protected: + void attachTableUnlocked(ContextPtr context, const String & table_name, const StoragePtr & table, const String & relative_table_path) TSA_REQUIRES(mutex) override; + private: struct CacheExpirationQueueElement { diff --git a/src/Databases/DatabaseMemory.cpp b/src/Databases/DatabaseMemory.cpp index 2a7a2ad8ccc..7be9d5e2cb7 100644 --- a/src/Databases/DatabaseMemory.cpp +++ b/src/Databases/DatabaseMemory.cpp @@ -33,13 +33,13 @@ DatabaseMemory::DatabaseMemory(const String & name_, ContextPtr context_) } void DatabaseMemory::createTable( - ContextPtr /*context*/, + ContextPtr local_context, const String & table_name, const StoragePtr & table, const ASTPtr & query) { std::lock_guard lock{mutex}; - attachTableUnlocked(table_name, table); + attachTableUnlocked(local_context, table_name, table, /*relative_table_path=*/ {}); /// Clean the query from temporary flags. ASTPtr query_to_store = query; @@ -56,7 +56,7 @@ void DatabaseMemory::createTable( } void DatabaseMemory::dropTable( - ContextPtr /*context*/, + ContextPtr local_context, const String & table_name, bool /*sync*/) { @@ -83,7 +83,7 @@ void DatabaseMemory::dropTable( catch (...) { std::lock_guard lock{mutex}; - attachTableUnlocked(table_name, table); + attachTableUnlocked(local_context, table_name, table, /*relative_table_path=*/ {}); throw; } diff --git a/src/Databases/DatabasesCommon.cpp b/src/Databases/DatabasesCommon.cpp index 7bc445c5b5d..512a7550050 100644 --- a/src/Databases/DatabasesCommon.cpp +++ b/src/Databases/DatabasesCommon.cpp @@ -7,6 +7,7 @@ #include #include #include +#include #include #include #include @@ -193,7 +194,7 @@ DatabaseWithOwnTablesBase::DatabaseWithOwnTablesBase(const String & name_, const bool DatabaseWithOwnTablesBase::isTableExist(const String & table_name, ContextPtr) const { std::lock_guard lock(mutex); - return tables.find(table_name) != tables.end(); + return tables.find(table_name) != tables.end() || lazy_tables.find(table_name) != lazy_tables.end(); } StoragePtr DatabaseWithOwnTablesBase::tryGetTable(const String & table_name, ContextPtr) const @@ -205,6 +206,9 @@ StoragePtr DatabaseWithOwnTablesBase::tryGetTable(const String & table_name, Con DatabaseTablesIteratorPtr DatabaseWithOwnTablesBase::getTablesIterator(ContextPtr, const FilterByNameFunction & filter_by_table_name) const { std::lock_guard lock(mutex); + + loadLazyTables(); + if (!filter_by_table_name) return std::make_unique(tables, database_name); @@ -250,13 +254,7 @@ StoragePtr DatabaseWithOwnTablesBase::detachTableUnlocked(const String & table_n return res; } -void DatabaseWithOwnTablesBase::attachTable(ContextPtr /* context_ */, const String & table_name, const StoragePtr & table, const String &) -{ - std::lock_guard lock(mutex); - attachTableUnlocked(table_name, table); -} - -void DatabaseWithOwnTablesBase::attachTableUnlocked(const String & table_name, const StoragePtr & table) +void DatabaseWithOwnTablesBase::attachTableUnlocked(ContextPtr, const String & name, const StoragePtr & table, const String &) { auto table_id = table->getStorageID(); if (table_id.database_name != database_name) @@ -269,7 +267,7 @@ void DatabaseWithOwnTablesBase::attachTableUnlocked(const String & table_name, c DatabaseCatalog::instance().addUUIDMapping(table_id.uuid, shared_from_this(), table); } - if (!tables.emplace(table_name, table).second) + if (!tables.emplace(name, table).second) { if (table_id.hasUUID()) DatabaseCatalog::instance().removeUUIDMapping(table_id.uuid); @@ -281,6 +279,12 @@ void DatabaseWithOwnTablesBase::attachTableUnlocked(const String & table_name, c table->is_detached = false; } +void DatabaseWithOwnTablesBase::registerLazyTableUnlocked(const String & table_name, LazyTableCreator table_creator, const String & relative_table_path) +{ + if (!lazy_tables.emplace(table_name, std::make_pair(relative_table_path, std::move(table_creator))).second) + throw Exception(ErrorCodes::TABLE_ALREADY_EXISTS, "Table {} already registered.", table_name); +} + void DatabaseWithOwnTablesBase::shutdown() { /// You can not hold a lock during shutdown. @@ -381,10 +385,45 @@ void DatabaseWithOwnTablesBase::createTableRestoredFromBackup(const ASTPtr & cre StoragePtr DatabaseWithOwnTablesBase::tryGetTableNoWait(const String & table_name) const { std::lock_guard lock(mutex); + auto it = tables.find(table_name); if (it != tables.end()) return it->second; + + const auto lazy_it = lazy_tables.find(table_name); + if (lazy_it != lazy_tables.end()) + { + LOG_DEBUG(log, "Attaching lazy table {}", backQuoteIfNeed(table_name)); + auto relative_table_path = lazy_it->second.first; + auto storage = lazy_it->second.second(); + lazy_tables.erase(lazy_it); + (const_cast(this))->attachTableUnlocked(Context::getGlobalContextInstance(), table_name, storage, relative_table_path); + + it = tables.find(table_name); + if (it != tables.end()) + return it->second; + } + return {}; } +void DatabaseWithOwnTablesBase::loadLazyTables() const +{ + if (lazy_tables.empty()) + return; + + ContextPtr global_context = Context::getGlobalContextInstance(); + while (!lazy_tables.empty()) + { + auto lazy_it = lazy_tables.begin(); + + const auto table_name = lazy_it->first; + LOG_DEBUG(log, "Attaching lazy table {}", backQuoteIfNeed(table_name)); + auto relative_table_path = lazy_it->second.first; + auto storage = lazy_it->second.second(); + lazy_tables.erase(lazy_it); + (const_cast(this))->attachTableUnlocked(global_context, table_name, storage, relative_table_path); + } +} + } diff --git a/src/Databases/DatabasesCommon.h b/src/Databases/DatabasesCommon.h index fc67596d3de..71704b95110 100644 --- a/src/Databases/DatabasesCommon.h +++ b/src/Databases/DatabasesCommon.h @@ -30,8 +30,6 @@ public: bool empty() const override; - void attachTable(ContextPtr context, const String & table_name, const StoragePtr & table, const String & relative_table_path) override; - StoragePtr detachTable(ContextPtr context, const String & table_name) override; DatabaseTablesIteratorPtr getTablesIterator(ContextPtr context, const FilterByNameFunction & filter_by_table_name) const override; @@ -45,14 +43,19 @@ public: protected: Tables tables TSA_GUARDED_BY(mutex); + /// Tables that are attached lazily + mutable LazyTables lazy_tables TSA_GUARDED_BY(mutex); Poco::Logger * log; DatabaseWithOwnTablesBase(const String & name_, const String & logger, ContextPtr context); - void attachTableUnlocked(const String & table_name, const StoragePtr & table) TSA_REQUIRES(mutex); + void attachTableUnlocked(ContextPtr context, const String & name, const StoragePtr & table, const String & relative_table_path) TSA_REQUIRES(mutex) override; + void registerLazyTableUnlocked(const String & table_name, LazyTableCreator table_creator, const String & relative_table_path) TSA_REQUIRES(mutex) override; StoragePtr detachTableUnlocked(const String & table_name) TSA_REQUIRES(mutex); StoragePtr getTableUnlocked(const String & table_name) const TSA_REQUIRES(mutex); StoragePtr tryGetTableNoWait(const String & table_name) const; + + void loadLazyTables() const TSA_REQUIRES(mutex); }; } diff --git a/src/Databases/IDatabase.cpp b/src/Databases/IDatabase.cpp index 09640d2f86e..ffd789dd531 100644 --- a/src/Databases/IDatabase.cpp +++ b/src/Databases/IDatabase.cpp @@ -46,4 +46,20 @@ void IDatabase::createTableRestoredFromBackup(const ASTPtr & create_table_query, backQuoteIfNeed(create_table_query->as().getTable())); } +/// Add a table to the database, but do not add it to the metadata. The database may not support this method. +/// +/// Note: ATTACH TABLE statement actually uses createTable method. +void IDatabase::attachTable(ContextPtr context, const String & name, const StoragePtr & table, const String & relative_table_path) /// NOLINT +{ + std::lock_guard lock(mutex); + attachTableUnlocked(context, name, table, relative_table_path); +} + +void IDatabase::registerLazyTable(ContextPtr, const String & table_name, LazyTableCreator table_creator, const String & relative_table_path) /// NOLINT +{ + std::lock_guard lock(mutex); + registerLazyTableUnlocked(table_name, std::move(table_creator), relative_table_path); +} + + } diff --git a/src/Databases/IDatabase.h b/src/Databases/IDatabase.h index e5afa0eb944..58e4c7e364b 100644 --- a/src/Databases/IDatabase.h +++ b/src/Databases/IDatabase.h @@ -125,7 +125,6 @@ public: using DatabaseTablesIteratorPtr = std::unique_ptr; - /** Database engine. * It is responsible for: * - initialization of set of known tables and dictionaries; @@ -138,6 +137,10 @@ using DatabaseTablesIteratorPtr = std::unique_ptr; class IDatabase : public std::enable_shared_from_this { public: + using LazyTableCreator = std::function; + /// Map{table_name, Pair{relative_table_path, LazyTableCreator}} + using LazyTables = std::map>; + IDatabase() = delete; explicit IDatabase(String database_name_) : database_name(std::move(database_name_)) {} @@ -269,11 +272,17 @@ public: /// Add a table to the database, but do not add it to the metadata. The database may not support this method. /// - /// Note: ATTACH TABLE statement actually uses createTable method. - virtual void attachTable(ContextPtr /* context */, const String & /*name*/, const StoragePtr & /*table*/, [[maybe_unused]] const String & relative_table_path = {}) /// NOLINT - { - throw Exception(ErrorCodes::NOT_IMPLEMENTED, "There is no ATTACH TABLE query for Database{}", getEngineName()); - } + /// @param relative_table_path - only for Atomic engine + /// + /// Note: + /// - ATTACH TABLE statement actually uses createTable method. + /// - Instead of overriding this method you should override attachTableUnlocked() + /// (This method is only for DatabasesOverlay to override) + virtual void attachTable(ContextPtr context, const String & name, const StoragePtr & table, const String & relative_table_path = {}); /// NOLINT + + /// Register tables lazily (attach will be done only when the table will be used) instead of attaching it. + /// This is needed to improve startup time of clickhouse-local. + virtual void registerLazyTable(ContextPtr context, const String & table_name, LazyTableCreator table_creator, const String & relative_table_path = {}); /// Forget about the table without deleting it, and return it. The database may not support this method. virtual StoragePtr detachTable(ContextPtr /* context */, const String & /*name*/) @@ -430,6 +439,16 @@ protected: return nullptr; } + virtual void attachTableUnlocked(ContextPtr /*context*/, const String & /*name*/, const StoragePtr & /*table*/, const String & /*relative_table_path*/ = {}) TSA_REQUIRES(mutex) /// NOLINT + { + throw Exception(ErrorCodes::NOT_IMPLEMENTED, "There is no ATTACH TABLE query for Database{}", getEngineName()); + } + + virtual void registerLazyTableUnlocked(const String & /* table_name */, LazyTableCreator /* table_creator */, const String & /* relative_table_path */) TSA_REQUIRES(mutex) /// NOLINT + { + throw Exception(ErrorCodes::NOT_IMPLEMENTED, "There lazy table initialization support for Database{}", getEngineName()); + } + mutable std::mutex mutex; String database_name TSA_GUARDED_BY(mutex); String comment TSA_GUARDED_BY(mutex); diff --git a/src/Databases/MySQL/DatabaseMaterializedMySQL.cpp b/src/Databases/MySQL/DatabaseMaterializedMySQL.cpp index a31e74cc7ae..965a3ac9965 100644 --- a/src/Databases/MySQL/DatabaseMaterializedMySQL.cpp +++ b/src/Databases/MySQL/DatabaseMaterializedMySQL.cpp @@ -101,10 +101,10 @@ void DatabaseMaterializedMySQL::dropTable(ContextPtr context_, const String & na DatabaseAtomic::dropTable(context_, name, sync); } -void DatabaseMaterializedMySQL::attachTable(ContextPtr context_, const String & name, const StoragePtr & table, const String & relative_table_path) +void DatabaseMaterializedMySQL::attachTableUnlocked(ContextPtr context_, const String & name, const StoragePtr & table, const String & relative_table_path) { checkIsInternalQuery(context_, "ATTACH TABLE"); - DatabaseAtomic::attachTable(context_, name, table, relative_table_path); + DatabaseAtomic::attachTableUnlocked(context_, name, table, relative_table_path); } StoragePtr DatabaseMaterializedMySQL::detachTable(ContextPtr context_, const String & name) diff --git a/src/Databases/MySQL/DatabaseMaterializedMySQL.h b/src/Databases/MySQL/DatabaseMaterializedMySQL.h index 895498723fd..0744b79257f 100644 --- a/src/Databases/MySQL/DatabaseMaterializedMySQL.h +++ b/src/Databases/MySQL/DatabaseMaterializedMySQL.h @@ -48,6 +48,8 @@ protected: LoadTaskPtr startup_mysql_database_task; + void attachTableUnlocked(ContextPtr context_, const String & name, const StoragePtr & table, const String & relative_table_path) TSA_REQUIRES(mutex) override; + public: String getEngineName() const override { return "MaterializedMySQL"; } @@ -58,8 +60,6 @@ public: void dropTable(ContextPtr context_, const String & name, bool sync) override; - void attachTable(ContextPtr context_, const String & name, const StoragePtr & table, const String & relative_table_path) override; - StoragePtr detachTable(ContextPtr context_, const String & name) override; void renameTable(ContextPtr context_, const String & name, IDatabase & to_database, const String & to_name, bool exchange, bool dictionary) override; diff --git a/src/Databases/MySQL/DatabaseMySQL.cpp b/src/Databases/MySQL/DatabaseMySQL.cpp index 7d2ed7a9662..42bceaa0e08 100644 --- a/src/Databases/MySQL/DatabaseMySQL.cpp +++ b/src/Databases/MySQL/DatabaseMySQL.cpp @@ -361,10 +361,8 @@ void DatabaseMySQL::cleanOutdatedTables() } } -void DatabaseMySQL::attachTable(ContextPtr /* context_ */, const String & table_name, const StoragePtr & storage, const String &) +void DatabaseMySQL::attachTableUnlocked(ContextPtr /* context_ */, const String & table_name, const StoragePtr & storage, const String &) { - std::lock_guard lock{mutex}; - if (!local_tables_cache.contains(table_name)) throw Exception(ErrorCodes::UNKNOWN_TABLE, "Cannot attach table {}.{} because it does not exist.", backQuoteIfNeed(database_name), backQuoteIfNeed(table_name)); diff --git a/src/Databases/MySQL/DatabaseMySQL.h b/src/Databases/MySQL/DatabaseMySQL.h index e5b1f434d2f..33bf26059c8 100644 --- a/src/Databases/MySQL/DatabaseMySQL.h +++ b/src/Databases/MySQL/DatabaseMySQL.h @@ -84,9 +84,9 @@ public: void dropTable(ContextPtr context, const String & table_name, bool sync) override; - void attachTable(ContextPtr context, const String & table_name, const StoragePtr & storage, const String & relative_table_path) override; - protected: + void attachTableUnlocked(ContextPtr context, const String & table_name, const StoragePtr & storage, const String & relative_table_path) TSA_REQUIRES(mutex) override; + ASTPtr getCreateTableQueryImpl(const String & name, ContextPtr context, bool throw_on_error) const override; private: diff --git a/src/Databases/PostgreSQL/DatabasePostgreSQL.cpp b/src/Databases/PostgreSQL/DatabasePostgreSQL.cpp index 24f04c16029..4f0f85bc787 100644 --- a/src/Databases/PostgreSQL/DatabasePostgreSQL.cpp +++ b/src/Databases/PostgreSQL/DatabasePostgreSQL.cpp @@ -216,10 +216,8 @@ StoragePtr DatabasePostgreSQL::fetchTable(const String & table_name, ContextPtr } -void DatabasePostgreSQL::attachTable(ContextPtr /* context_ */, const String & table_name, const StoragePtr & storage, const String &) +void DatabasePostgreSQL::attachTableUnlocked(ContextPtr /* context_ */, const String & table_name, const StoragePtr & storage, const String &) { - std::lock_guard lock{mutex}; - if (!checkPostgresTable(table_name)) throw Exception(ErrorCodes::UNKNOWN_TABLE, "Cannot attach PostgreSQL table {} because it does not exist in PostgreSQL (database: {})", diff --git a/src/Databases/PostgreSQL/DatabasePostgreSQL.h b/src/Databases/PostgreSQL/DatabasePostgreSQL.h index d731e06649b..30d0070b2d7 100644 --- a/src/Databases/PostgreSQL/DatabasePostgreSQL.h +++ b/src/Databases/PostgreSQL/DatabasePostgreSQL.h @@ -54,13 +54,14 @@ public: void createTable(ContextPtr, const String & table_name, const StoragePtr & storage, const ASTPtr & create_query) override; void dropTable(ContextPtr, const String & table_name, bool sync) override; - void attachTable(ContextPtr context, const String & table_name, const StoragePtr & storage, const String & relative_table_path) override; StoragePtr detachTable(ContextPtr context, const String & table_name) override; void drop(ContextPtr /*context*/) override; void shutdown() override; protected: + void attachTableUnlocked(ContextPtr context, const String & table_name, const StoragePtr & storage, const String & relative_table_path) TSA_REQUIRES(mutex) override; + ASTPtr getCreateTableQueryImpl(const String & table_name, ContextPtr context, bool throw_on_error) const override; private: diff --git a/src/Disks/IO/createReadBufferFromFileBase.cpp b/src/Disks/IO/createReadBufferFromFileBase.cpp index 80dbc8df988..236dd43e9ee 100644 --- a/src/Disks/IO/createReadBufferFromFileBase.cpp +++ b/src/Disks/IO/createReadBufferFromFileBase.cpp @@ -101,12 +101,16 @@ std::unique_ptr createReadBufferFromFileBase( else if (settings.local_fs_method == LocalFSReadMethod::io_uring) { #if USE_LIBURING - static std::shared_ptr reader = std::make_shared(512); - if (!reader->isSupported()) + auto global_context = Context::getGlobalContextInstance(); + if (!global_context) + throw Exception(ErrorCodes::LOGICAL_ERROR, "Cannot obtain io_uring reader (global context not initialized)"); + + auto & reader = global_context->getIOURingReader(); + if (!reader.isSupported()) throw Exception(ErrorCodes::UNSUPPORTED_METHOD, "io_uring is not supported by this system"); res = std::make_unique( - *reader, + reader, settings.priority, filename, buffer_size, diff --git a/src/Functions/FunctionsAES.cpp b/src/Functions/FunctionsAES.cpp index 2911d115e35..d1fcd192e6f 100644 --- a/src/Functions/FunctionsAES.cpp +++ b/src/Functions/FunctionsAES.cpp @@ -15,14 +15,13 @@ namespace ErrorCodes { extern const int OPENSSL_ERROR; } -} + namespace OpenSSLDetails { void onError(std::string error_message) { - error_message += ". OpenSSL error code: " + std::to_string(ERR_get_error()); - throw DB::Exception::createDeprecated(error_message, DB::ErrorCodes::OPENSSL_ERROR); + throw Exception(ErrorCodes::OPENSSL_ERROR, "{}. OpenSSL error code: {}", error_message, ERR_get_error()); } StringRef foldEncryptionKeyInMySQLCompatitableMode(size_t cipher_key_size, StringRef key, std::array & folded_key) @@ -48,4 +47,6 @@ const EVP_CIPHER * getCipherByName(StringRef cipher_name) } +} + #endif diff --git a/src/Functions/FunctionsAES.h b/src/Functions/FunctionsAES.h index c748be5b9b8..4792c997f51 100644 --- a/src/Functions/FunctionsAES.h +++ b/src/Functions/FunctionsAES.h @@ -25,13 +25,14 @@ #include + namespace DB { namespace ErrorCodes { extern const int BAD_ARGUMENTS; } -} + namespace OpenSSLDetails { @@ -60,7 +61,7 @@ struct KeyHolder inline StringRef setKey(size_t cipher_key_size, StringRef key) const { if (key.size != cipher_key_size) - throw DB::Exception(DB::ErrorCodes::BAD_ARGUMENTS, "Invalid key size: {} expected {}", key.size, cipher_key_size); + throw Exception(ErrorCodes::BAD_ARGUMENTS, "Invalid key size: {} expected {}", key.size, cipher_key_size); return key; } @@ -72,7 +73,7 @@ struct KeyHolder inline StringRef setKey(size_t cipher_key_size, StringRef key) { if (key.size < cipher_key_size) - throw DB::Exception(DB::ErrorCodes::BAD_ARGUMENTS, "Invalid key size: {} expected {}", key.size, cipher_key_size); + throw Exception(ErrorCodes::BAD_ARGUMENTS, "Invalid key size: {} expected {}", key.size, cipher_key_size); // MySQL does something fancy with the keys that are too long, // ruining compatibility with OpenSSL and not improving security. @@ -118,7 +119,7 @@ inline void validateCipherMode(const EVP_CIPHER * evp_cipher) } } - throw DB::Exception(DB::ErrorCodes::BAD_ARGUMENTS, "Unsupported cipher mode"); + throw Exception(ErrorCodes::BAD_ARGUMENTS, "Unsupported cipher mode"); } template @@ -127,13 +128,11 @@ inline void validateIV(StringRef iv_value, const size_t cipher_iv_size) // In MySQL mode we don't care if IV is longer than expected, only if shorter. if ((mode == CipherMode::MySQLCompatibility && iv_value.size != 0 && iv_value.size < cipher_iv_size) || (mode == CipherMode::OpenSSLCompatibility && iv_value.size != 0 && iv_value.size != cipher_iv_size)) - throw DB::Exception(DB::ErrorCodes::BAD_ARGUMENTS, "Invalid IV size: {} expected {}", iv_value.size, cipher_iv_size); + throw Exception(ErrorCodes::BAD_ARGUMENTS, "Invalid IV size: {} expected {}", iv_value.size, cipher_iv_size); } } -namespace DB -{ template class FunctionEncrypt : public IFunction { @@ -313,12 +312,12 @@ private: // in GCM mode IV can be of arbitrary size (>0), IV is optional for other modes. if (mode == CipherMode::RFC5116_AEAD_AES_GCM && iv_value.size == 0) { - throw Exception(DB::ErrorCodes::BAD_ARGUMENTS, "Invalid IV size {} != expected size {}", iv_value.size, iv_size); + throw Exception(ErrorCodes::BAD_ARGUMENTS, "Invalid IV size {} != expected size {}", iv_value.size, iv_size); } if (mode != CipherMode::RFC5116_AEAD_AES_GCM && key_value.size != key_size) { - throw Exception(DB::ErrorCodes::BAD_ARGUMENTS, "Invalid key size {} != expected size {}", key_value.size, key_size); + throw Exception(ErrorCodes::BAD_ARGUMENTS, "Invalid key size {} != expected size {}", key_value.size, key_size); } } @@ -608,12 +607,12 @@ private: // in GCM mode IV can be of arbitrary size (>0), for other modes IV is optional. if (mode == CipherMode::RFC5116_AEAD_AES_GCM && iv_value.size == 0) { - throw Exception(DB::ErrorCodes::BAD_ARGUMENTS, "Invalid IV size {} != expected size {}", iv_value.size, iv_size); + throw Exception(ErrorCodes::BAD_ARGUMENTS, "Invalid IV size {} != expected size {}", iv_value.size, iv_size); } if (key_value.size != key_size) { - throw Exception(DB::ErrorCodes::BAD_ARGUMENTS, "Invalid key size {} != expected size {}", key_value.size, key_size); + throw Exception(ErrorCodes::BAD_ARGUMENTS, "Invalid key size {} != expected size {}", key_value.size, key_size); } } diff --git a/src/Functions/FunctionsHashing.h b/src/Functions/FunctionsHashing.h index 345b9a11e0d..9468bc259e3 100644 --- a/src/Functions/FunctionsHashing.h +++ b/src/Functions/FunctionsHashing.h @@ -28,6 +28,11 @@ # include # include # include +#if USE_BORINGSSL +# include +#else +# include +#endif #endif #include @@ -318,6 +323,25 @@ struct SHA512Impl SHA512_Final(out_char_data, &ctx); } }; + +struct SHA512Impl256 +{ + static constexpr auto name = "SHA512_256"; + enum { length = 32 }; + + static void apply(const char * begin, const size_t size, unsigned char * out_char_data) + { + /// Here, we use the EVP interface that is common to both BoringSSL and OpenSSL. Though BoringSSL is the default + /// SSL library that we use, for S390X architecture only OpenSSL is supported. But the SHA512-256, SHA512_256_Init, + /// SHA512_256_Update, SHA512_256_Final methods to calculate hash (similar to the other SHA functions) aren't available + /// in the current version of OpenSSL that we use which necessitates the use of the EVP interface. + auto md_ctx = EVP_MD_CTX_create(); + EVP_DigestInit_ex(md_ctx, EVP_sha512_256(), nullptr /*engine*/); + EVP_DigestUpdate(md_ctx, begin, size); + EVP_DigestFinal_ex(md_ctx, out_char_data, nullptr /*size*/); + EVP_MD_CTX_destroy(md_ctx); + } +}; #endif struct SipHash64Impl @@ -1801,6 +1825,7 @@ using FunctionSHA224 = FunctionStringHashFixedString; using FunctionSHA256 = FunctionStringHashFixedString; using FunctionSHA384 = FunctionStringHashFixedString; using FunctionSHA512 = FunctionStringHashFixedString; +using FunctionSHA512_256 = FunctionStringHashFixedString; #endif using FunctionSipHash128 = FunctionAnyHash; using FunctionSipHash128Keyed = FunctionAnyHash; diff --git a/src/Functions/FunctionsHashingSSL.cpp b/src/Functions/FunctionsHashingSSL.cpp index 2eb0ed88f28..b716a11f9c3 100644 --- a/src/Functions/FunctionsHashingSSL.cpp +++ b/src/Functions/FunctionsHashingSSL.cpp @@ -22,6 +22,22 @@ REGISTER_FUNCTION(HashingSSL) factory.registerFunction(); factory.registerFunction(); factory.registerFunction(); + factory.registerFunction(FunctionDocumentation{ + .description = R"(Calculates the SHA512_256 hash of the given string.)", + .syntax = "SELECT SHA512_256(s);", + .arguments = {{"s", "The input [String](../../sql-reference/data-types/string.md)."}}, + .returned_value + = "The SHA512_256 hash of the given input string returned as a [FixedString](../../sql-reference/data-types/fixedstring.md).", + .examples + = {{"", + "SELECT HEX(SHA512_256('abc'));", + R"( +┌─hex(SHA512_256('abc'))───────────────────────────────────────────┐ +│ 53048E2681941EF99B2E29B76B4C7DABE4C2D0C634FC6D46E0E2F13107E7AF23 │ +└──────────────────────────────────────────────────────────────────┘ + )" + }} + }); } } diff --git a/src/Functions/aes_decrypt_mysql.cpp b/src/Functions/aes_decrypt_mysql.cpp index bb8ef3809d1..8043fa422ad 100644 --- a/src/Functions/aes_decrypt_mysql.cpp +++ b/src/Functions/aes_decrypt_mysql.cpp @@ -5,6 +5,10 @@ #include #include + +namespace DB +{ + namespace { @@ -17,9 +21,6 @@ struct DecryptMySQLModeImpl } -namespace DB -{ - REGISTER_FUNCTION(AESDecryptMysql) { factory.registerFunction>(); diff --git a/src/Functions/aes_encrypt_mysql.cpp b/src/Functions/aes_encrypt_mysql.cpp index 0dcb4108770..fb120151c25 100644 --- a/src/Functions/aes_encrypt_mysql.cpp +++ b/src/Functions/aes_encrypt_mysql.cpp @@ -5,6 +5,9 @@ #include #include +namespace DB +{ + namespace { @@ -16,9 +19,6 @@ struct EncryptMySQLModeImpl } -namespace DB -{ - REGISTER_FUNCTION(AESEncryptMysql) { factory.registerFunction>(); diff --git a/src/Functions/date_trunc.cpp b/src/Functions/date_trunc.cpp index 414512fc4f8..c3903fef137 100644 --- a/src/Functions/date_trunc.cpp +++ b/src/Functions/date_trunc.cpp @@ -47,7 +47,7 @@ public: throw Exception(ErrorCodes::ILLEGAL_TYPE_OF_ARGUMENT, "First argument for function {} must be constant string: " "name of datepart", getName()); - datepart_param = datepart_column->getValue(); + datepart_param = Poco::toLower(datepart_column->getValue()); if (datepart_param.empty()) throw Exception(ErrorCodes::BAD_ARGUMENTS, "First argument (name of datepart) for function {} cannot be empty", getName()); diff --git a/src/Functions/decrypt.cpp b/src/Functions/decrypt.cpp index 1a2d5e2b621..6ad58b959e9 100644 --- a/src/Functions/decrypt.cpp +++ b/src/Functions/decrypt.cpp @@ -5,6 +5,9 @@ #include #include +namespace DB +{ + namespace { @@ -17,9 +20,6 @@ struct DecryptImpl } -namespace DB -{ - REGISTER_FUNCTION(Decrypt) { factory.registerFunction>(); diff --git a/src/Functions/encrypt.cpp b/src/Functions/encrypt.cpp index 38feafbea19..d76148f2176 100644 --- a/src/Functions/encrypt.cpp +++ b/src/Functions/encrypt.cpp @@ -5,6 +5,9 @@ #include #include +namespace DB +{ + namespace { @@ -16,9 +19,6 @@ struct EncryptImpl } -namespace DB -{ - REGISTER_FUNCTION(Encrypt) { factory.registerFunction>(); diff --git a/src/Functions/tryDecrypt.cpp b/src/Functions/tryDecrypt.cpp index eccccb1e4a4..90d10103dd1 100644 --- a/src/Functions/tryDecrypt.cpp +++ b/src/Functions/tryDecrypt.cpp @@ -3,8 +3,12 @@ #if USE_SSL -# include -# include +#include +#include + + +namespace DB +{ namespace { @@ -18,9 +22,6 @@ struct TryDecryptImpl } -namespace DB -{ - REGISTER_FUNCTION(TryDecrypt) { factory.registerFunction>(FunctionDocumentation{ diff --git a/src/IO/FileEncryptionCommon.cpp b/src/IO/FileEncryptionCommon.cpp index 6f4db4975f7..c3348ef6068 100644 --- a/src/IO/FileEncryptionCommon.cpp +++ b/src/IO/FileEncryptionCommon.cpp @@ -97,7 +97,7 @@ namespace uint8_t * ciphertext = reinterpret_cast(out.position()); int ciphertext_size = 0; if (!EVP_EncryptUpdate(evp_ctx, ciphertext, &ciphertext_size, &in[in_size], static_cast(part_size))) - throw Exception(ErrorCodes::DATA_ENCRYPTION_ERROR, "Failed to encrypt"); + throw Exception(ErrorCodes::DATA_ENCRYPTION_ERROR, "Failed to encrypt: {}", ERR_get_error()); in_size += part_size; if (ciphertext_size) @@ -120,7 +120,7 @@ namespace uint8_t ciphertext[kBlockSize]; int ciphertext_size = 0; if (!EVP_EncryptUpdate(evp_ctx, ciphertext, &ciphertext_size, padded_data, safe_cast(padded_data_size))) - throw Exception(ErrorCodes::DATA_ENCRYPTION_ERROR, "Failed to encrypt"); + throw Exception(ErrorCodes::DATA_ENCRYPTION_ERROR, "Failed to encrypt: {}", ERR_get_error()); if (!ciphertext_size) return 0; @@ -140,7 +140,7 @@ namespace int ciphertext_size = 0; if (!EVP_EncryptFinal_ex(evp_ctx, ciphertext, &ciphertext_size)) - throw Exception(ErrorCodes::DATA_ENCRYPTION_ERROR, "Failed to finalize encrypting"); + throw Exception(ErrorCodes::DATA_ENCRYPTION_ERROR, "Failed to finalize encrypting: {}", ERR_get_error()); if (ciphertext_size) out.write(reinterpret_cast(ciphertext), ciphertext_size); return ciphertext_size; @@ -152,7 +152,7 @@ namespace uint8_t * plaintext = reinterpret_cast(out); int plaintext_size = 0; if (!EVP_DecryptUpdate(evp_ctx, plaintext, &plaintext_size, in, safe_cast(size))) - throw Exception(ErrorCodes::DATA_ENCRYPTION_ERROR, "Failed to decrypt"); + throw Exception(ErrorCodes::DATA_ENCRYPTION_ERROR, "Failed to decrypt: {}", ERR_get_error()); return plaintext_size; } @@ -165,7 +165,7 @@ namespace uint8_t plaintext[kBlockSize]; int plaintext_size = 0; if (!EVP_DecryptUpdate(evp_ctx, plaintext, &plaintext_size, padded_data, safe_cast(padded_data_size))) - throw Exception(ErrorCodes::DATA_ENCRYPTION_ERROR, "Failed to decrypt"); + throw Exception(ErrorCodes::DATA_ENCRYPTION_ERROR, "Failed to decrypt: {}", ERR_get_error()); if (!plaintext_size) return 0; @@ -184,7 +184,7 @@ namespace uint8_t plaintext[kBlockSize]; int plaintext_size = 0; if (!EVP_DecryptFinal_ex(evp_ctx, plaintext, &plaintext_size)) - throw Exception(ErrorCodes::DATA_ENCRYPTION_ERROR, "Failed to finalize decrypting"); + throw Exception(ErrorCodes::DATA_ENCRYPTION_ERROR, "Failed to finalize decrypting: {}", ERR_get_error()); if (plaintext_size) memcpy(out, plaintext, plaintext_size); return plaintext_size; @@ -291,11 +291,11 @@ void Encryptor::encrypt(const char * data, size_t size, WriteBuffer & out) auto * evp_ctx = evp_ctx_ptr.get(); if (!EVP_EncryptInit_ex(evp_ctx, evp_cipher, nullptr, nullptr, nullptr)) - throw Exception(ErrorCodes::DATA_ENCRYPTION_ERROR, "Failed to initialize encryption context with cipher"); + throw Exception(ErrorCodes::DATA_ENCRYPTION_ERROR, "Failed to initialize encryption context with cipher: {}", ERR_get_error()); if (!EVP_EncryptInit_ex(evp_ctx, nullptr, nullptr, reinterpret_cast(key.c_str()), reinterpret_cast(current_iv.c_str()))) - throw Exception(ErrorCodes::DATA_ENCRYPTION_ERROR, "Failed to set key and IV for encryption"); + throw Exception(ErrorCodes::DATA_ENCRYPTION_ERROR, "Failed to set key and IV for encryption: {}", ERR_get_error()); size_t in_size = 0; size_t out_size = 0; @@ -320,7 +320,7 @@ void Encryptor::encrypt(const char * data, size_t size, WriteBuffer & out) out_size += encryptFinal(evp_ctx, out); if (out_size != in_size) - throw Exception(ErrorCodes::DATA_ENCRYPTION_ERROR, "Only part of the data was encrypted"); + throw Exception(ErrorCodes::DATA_ENCRYPTION_ERROR, "Only part of the data was encrypted: {} out of {} bytes", out_size, in_size); offset += in_size; } @@ -335,11 +335,11 @@ void Encryptor::decrypt(const char * data, size_t size, char * out) auto * evp_ctx = evp_ctx_ptr.get(); if (!EVP_DecryptInit_ex(evp_ctx, evp_cipher, nullptr, nullptr, nullptr)) - throw Exception(ErrorCodes::DATA_ENCRYPTION_ERROR, "Failed to initialize decryption context with cipher"); + throw Exception(ErrorCodes::DATA_ENCRYPTION_ERROR, "Failed to initialize decryption context with cipher: {}", ERR_get_error()); if (!EVP_DecryptInit_ex(evp_ctx, nullptr, nullptr, reinterpret_cast(key.c_str()), reinterpret_cast(current_iv.c_str()))) - throw Exception(ErrorCodes::DATA_ENCRYPTION_ERROR, "Failed to set key and IV for decryption"); + throw Exception(ErrorCodes::DATA_ENCRYPTION_ERROR, "Failed to set key and IV for decryption: {}", ERR_get_error()); size_t in_size = 0; size_t out_size = 0; @@ -364,7 +364,7 @@ void Encryptor::decrypt(const char * data, size_t size, char * out) out_size += decryptFinal(evp_ctx, &out[out_size]); if (out_size != in_size) - throw Exception(ErrorCodes::DATA_ENCRYPTION_ERROR, "Only part of the data was decrypted"); + throw Exception(ErrorCodes::DATA_ENCRYPTION_ERROR, "Only part of the data was decrypted: {} out of {} bytes", out_size, in_size); offset += in_size; } diff --git a/src/IO/ReadBufferFromFileDescriptor.cpp b/src/IO/ReadBufferFromFileDescriptor.cpp index 6c0c1681a4c..5a67dc7528c 100644 --- a/src/IO/ReadBufferFromFileDescriptor.cpp +++ b/src/IO/ReadBufferFromFileDescriptor.cpp @@ -39,7 +39,6 @@ namespace ErrorCodes extern const int CANNOT_READ_FROM_FILE_DESCRIPTOR; extern const int ARGUMENT_OUT_OF_BOUND; extern const int CANNOT_SEEK_THROUGH_FILE; - extern const int CANNOT_SELECT; extern const int CANNOT_ADVISE; } @@ -249,24 +248,6 @@ void ReadBufferFromFileDescriptor::rewind() file_offset_of_buffer_end = 0; } - -/// Assuming file descriptor supports 'select', check that we have data to read or wait until timeout. -bool ReadBufferFromFileDescriptor::poll(size_t timeout_microseconds) const -{ - fd_set fds; - FD_ZERO(&fds); - FD_SET(fd, &fds); - timeval timeout = { time_t(timeout_microseconds / 1000000), suseconds_t(timeout_microseconds % 1000000) }; - - int res = select(1, &fds, nullptr, nullptr, &timeout); - - if (-1 == res) - throwFromErrno("Cannot select", ErrorCodes::CANNOT_SELECT); - - return res > 0; -} - - size_t ReadBufferFromFileDescriptor::getFileSize() { return getSizeFromFileDescriptor(fd, getFileName()); diff --git a/src/IO/ReadBufferFromFileDescriptor.h b/src/IO/ReadBufferFromFileDescriptor.h index 64340770cf2..4762998c67b 100644 --- a/src/IO/ReadBufferFromFileDescriptor.h +++ b/src/IO/ReadBufferFromFileDescriptor.h @@ -75,10 +75,6 @@ public: size_t readBigAt(char * to, size_t n, size_t offset, const std::function &) override; bool supportsReadAt() override { return use_pread; } - -private: - /// Assuming file descriptor supports 'select', check that we have data to read or wait until timeout. - bool poll(size_t timeout_microseconds) const; }; diff --git a/src/Interpreters/Context.cpp b/src/Interpreters/Context.cpp index ea193e819ad..aa4ef4155a6 100644 --- a/src/Interpreters/Context.cpp +++ b/src/Interpreters/Context.cpp @@ -35,6 +35,7 @@ #include #include #include +#include #include #include #include @@ -309,6 +310,11 @@ struct ContextSharedPart : boost::noncopyable mutable OnceFlag threadpool_writer_initialized; mutable std::unique_ptr threadpool_writer; +#if USE_LIBURING + mutable OnceFlag io_uring_reader_initialized; + mutable std::unique_ptr io_uring_reader; +#endif + mutable ThrottlerPtr replicated_fetches_throttler; /// A server-wide throttler for replicated fetches mutable ThrottlerPtr replicated_sends_throttler; /// A server-wide throttler for replicated sends @@ -4811,7 +4817,7 @@ void Context::initializeBackgroundExecutorsIfNeeded() shared->are_background_executors_initialized = true; } -bool Context::areBackgroundExecutorsInitialized() +bool Context::areBackgroundExecutorsInitialized() const { SharedLockGuard lock(shared->background_executors_mutex); return shared->are_background_executors_initialized; @@ -4861,6 +4867,17 @@ IAsynchronousReader & Context::getThreadPoolReader(FilesystemReaderType type) co } } +#if USE_LIBURING +IOUringReader & Context::getIOURingReader() const +{ + callOnce(shared->io_uring_reader_initialized, [&] { + shared->io_uring_reader = std::make_unique(512); + }); + + return *shared->io_uring_reader; +} +#endif + ThreadPool & Context::getThreadPoolWriter() const { callOnce(shared->threadpool_writer_initialized, [&] { diff --git a/src/Interpreters/Context.h b/src/Interpreters/Context.h index 1e02e30c111..c798837a6f5 100644 --- a/src/Interpreters/Context.h +++ b/src/Interpreters/Context.h @@ -109,6 +109,7 @@ class AsynchronousInsertLog; class BackupLog; class BlobStorageLog; class IAsynchronousReader; +class IOUringReader; struct MergeTreeSettings; struct InitialAllRangesAnnouncement; struct ParallelReadRequest; @@ -1202,7 +1203,7 @@ public: /// Background executors related methods void initializeBackgroundExecutorsIfNeeded(); - bool areBackgroundExecutorsInitialized(); + bool areBackgroundExecutorsInitialized() const; MergeMutateBackgroundExecutorPtr getMergeMutateExecutor() const; OrdinaryBackgroundExecutorPtr getMovesExecutor() const; @@ -1210,6 +1211,9 @@ public: OrdinaryBackgroundExecutorPtr getCommonExecutor() const; IAsynchronousReader & getThreadPoolReader(FilesystemReaderType type) const; +#if USE_LIBURING + IOUringReader & getIOURingReader() const; +#endif std::shared_ptr getAsyncReadCounters() const; diff --git a/src/Interpreters/ServerAsynchronousMetrics.cpp b/src/Interpreters/ServerAsynchronousMetrics.cpp index 84d31bae13f..8cf7dc39d97 100644 --- a/src/Interpreters/ServerAsynchronousMetrics.cpp +++ b/src/Interpreters/ServerAsynchronousMetrics.cpp @@ -54,8 +54,8 @@ ServerAsynchronousMetrics::ServerAsynchronousMetrics( int update_period_seconds, int heavy_metrics_update_period_seconds, const ProtocolServerMetricsFunc & protocol_server_metrics_func_) - : AsynchronousMetrics(update_period_seconds, protocol_server_metrics_func_) - , WithContext(global_context_) + : WithContext(global_context_) + , AsynchronousMetrics(update_period_seconds, protocol_server_metrics_func_) , heavy_metric_update_period(heavy_metrics_update_period_seconds) { /// sanity check @@ -63,6 +63,12 @@ ServerAsynchronousMetrics::ServerAsynchronousMetrics( throw Exception(ErrorCodes::INVALID_SETTING_VALUE, "Setting asynchronous_metrics_update_period_s and asynchronous_heavy_metrics_update_period_s must not be zero"); } +ServerAsynchronousMetrics::~ServerAsynchronousMetrics() +{ + /// NOTE: stop() from base class is not enough, since this leads to leak on vptr + stop(); +} + void ServerAsynchronousMetrics::updateImpl(AsynchronousMetricValues & new_values, TimePoint update_time, TimePoint current_time) { if (auto mark_cache = getContext()->getMarkCache()) diff --git a/src/Interpreters/ServerAsynchronousMetrics.h b/src/Interpreters/ServerAsynchronousMetrics.h index 8243699a111..a579d12de2c 100644 --- a/src/Interpreters/ServerAsynchronousMetrics.h +++ b/src/Interpreters/ServerAsynchronousMetrics.h @@ -7,7 +7,7 @@ namespace DB { -class ServerAsynchronousMetrics : public AsynchronousMetrics, WithContext +class ServerAsynchronousMetrics : WithContext, public AsynchronousMetrics { public: ServerAsynchronousMetrics( @@ -15,6 +15,8 @@ public: int update_period_seconds, int heavy_metrics_update_period_seconds, const ProtocolServerMetricsFunc & protocol_server_metrics_func_); + ~ServerAsynchronousMetrics() override; + private: void updateImpl(AsynchronousMetricValues & new_values, TimePoint update_time, TimePoint current_time) override; void logImpl(AsynchronousMetricValues & new_values) override; diff --git a/src/Planner/PlannerJoinTree.cpp b/src/Planner/PlannerJoinTree.cpp index 95ea260e727..abcf971b832 100644 --- a/src/Planner/PlannerJoinTree.cpp +++ b/src/Planner/PlannerJoinTree.cpp @@ -988,18 +988,7 @@ JoinTreeQueryPlan buildQueryPlanForJoinNode(const QueryTreeNodePtr & join_table_ if (join_strictness == JoinStrictness::All || join_strictness == JoinStrictness::Semi || join_strictness == JoinStrictness::Anti) join_constant = tryExtractConstantFromJoinNode(join_table_expression); - if (join_constant) - { - /** If there is JOIN with always true constant, we transform it to cross. - * If there is JOIN with always false constant, we do not process JOIN keys. - * It is expected by join algorithm to handle such case. - * - * Example: SELECT * FROM test_table AS t1 INNER JOIN test_table AS t2 ON 1; - */ - if (*join_constant) - join_kind = JoinKind::Cross; - } - else if (join_node.isOnJoinExpression()) + if (!join_constant && join_node.isOnJoinExpression()) { join_clauses_and_actions = buildJoinClausesAndActions(left_plan_output_columns, right_plan_output_columns, @@ -1079,7 +1068,6 @@ JoinTreeQueryPlan buildQueryPlanForJoinNode(const QueryTreeNodePtr & join_table_ const auto & query_context = planner_context->getQueryContext(); const auto & settings = query_context->getSettingsRef(); - bool join_use_nulls = settings.join_use_nulls; auto to_nullable_function = FunctionFactory::instance().get("toNullable", query_context); auto join_cast_plan_columns_to_nullable = [&](QueryPlan & plan_to_add_cast) @@ -1105,7 +1093,7 @@ JoinTreeQueryPlan buildQueryPlanForJoinNode(const QueryTreeNodePtr & join_table_ plan_to_add_cast.addStep(std::move(cast_join_columns_step)); }; - if (join_use_nulls) + if (settings.join_use_nulls) { if (isFull(join_kind)) { @@ -1124,6 +1112,18 @@ JoinTreeQueryPlan buildQueryPlanForJoinNode(const QueryTreeNodePtr & join_table_ auto table_join = std::make_shared(settings, query_context->getGlobalTemporaryVolume()); table_join->getTableJoin() = join_node.toASTTableJoin()->as(); + + if (join_constant) + { + /** If there is JOIN with always true constant, we transform it to cross. + * If there is JOIN with always false constant, we do not process JOIN keys. + * It is expected by join algorithm to handle such case. + * + * Example: SELECT * FROM test_table AS t1 INNER JOIN test_table AS t2 ON 1; + */ + if (*join_constant) + join_kind = JoinKind::Cross; + } table_join->getTableJoin().kind = join_kind; if (join_kind == JoinKind::Comma) diff --git a/src/Storages/AlterCommands.cpp b/src/Storages/AlterCommands.cpp index d19f92ae767..6f93cb3c370 100644 --- a/src/Storages/AlterCommands.cpp +++ b/src/Storages/AlterCommands.cpp @@ -707,7 +707,7 @@ void AlterCommand::apply(StorageInMemoryMetadata & metadata, ContextPtr context) } else if (type == MODIFY_TTL) { - metadata.table_ttl = TTLTableDescription::getTTLForTableFromAST(ttl, metadata.columns, context, metadata.primary_key); + metadata.table_ttl = TTLTableDescription::getTTLForTableFromAST(ttl, metadata.columns, context, metadata.primary_key, context->getSettingsRef().allow_suspicious_ttl_expressions); } else if (type == REMOVE_TTL) { @@ -1136,13 +1136,13 @@ void AlterCommands::apply(StorageInMemoryMetadata & metadata, ContextPtr context metadata_copy.column_ttls_by_name.clear(); for (const auto & [name, ast] : column_ttl_asts) { - auto new_ttl_entry = TTLDescription::getTTLFromAST(ast, metadata_copy.columns, context, metadata_copy.primary_key); + auto new_ttl_entry = TTLDescription::getTTLFromAST(ast, metadata_copy.columns, context, metadata_copy.primary_key, context->getSettingsRef().allow_suspicious_ttl_expressions); metadata_copy.column_ttls_by_name[name] = new_ttl_entry; } if (metadata_copy.table_ttl.definition_ast != nullptr) metadata_copy.table_ttl = TTLTableDescription::getTTLForTableFromAST( - metadata_copy.table_ttl.definition_ast, metadata_copy.columns, context, metadata_copy.primary_key); + metadata_copy.table_ttl.definition_ast, metadata_copy.columns, context, metadata_copy.primary_key, context->getSettingsRef().allow_suspicious_ttl_expressions); metadata = std::move(metadata_copy); } diff --git a/src/Storages/Distributed/DistributedAsyncInsertBatch.cpp b/src/Storages/Distributed/DistributedAsyncInsertBatch.cpp index b82cf1d7ffc..97268cf1389 100644 --- a/src/Storages/Distributed/DistributedAsyncInsertBatch.cpp +++ b/src/Storages/Distributed/DistributedAsyncInsertBatch.cpp @@ -5,6 +5,7 @@ #include #include #include +#include #include #include @@ -163,6 +164,22 @@ void DistributedAsyncInsertBatch::deserialize() readText(in); } +bool DistributedAsyncInsertBatch::valid() +{ + chassert(!files.empty()); + + bool res = true; + for (const auto & file : files) + { + if (!fs::exists(file)) + { + LOG_WARNING(parent.log, "File {} does not exists, likely due abnormal shutdown", file); + res = false; + } + } + return res; +} + void DistributedAsyncInsertBatch::writeText(WriteBuffer & out) { for (const auto & file : files) @@ -201,14 +218,6 @@ void DistributedAsyncInsertBatch::sendBatch() { for (const auto & file : files) { - /// In case of recovery it is possible that some of files will be - /// missing, if server had been restarted abnormally - if (recovered && !fs::exists(file)) - { - LOG_WARNING(parent.log, "File {} does not exists, likely due abnormal shutdown", file); - continue; - } - ReadBufferFromFile in(file); const auto & distributed_header = DistributedAsyncInsertHeader::read(in, parent.log); diff --git a/src/Storages/Distributed/DistributedAsyncInsertBatch.h b/src/Storages/Distributed/DistributedAsyncInsertBatch.h index 867a0de89fa..db96634d6f1 100644 --- a/src/Storages/Distributed/DistributedAsyncInsertBatch.h +++ b/src/Storages/Distributed/DistributedAsyncInsertBatch.h @@ -18,9 +18,16 @@ public: bool isEnoughSize() const; void send(); + /// Write batch to current_batch.txt void serialize(); + + /// Read batch from current_batch.txt void deserialize(); + /// Does all required files exists? + /// (The only way variant when it is valid is during restoring batch from disk). + bool valid(); + size_t total_rows = 0; size_t total_bytes = 0; std::vector files; diff --git a/src/Storages/Distributed/DistributedAsyncInsertDirectoryQueue.cpp b/src/Storages/Distributed/DistributedAsyncInsertDirectoryQueue.cpp index 66dcb9ad05f..26fa489a63d 100644 --- a/src/Storages/Distributed/DistributedAsyncInsertDirectoryQueue.cpp +++ b/src/Storages/Distributed/DistributedAsyncInsertDirectoryQueue.cpp @@ -16,13 +16,14 @@ #include #include #include -#include +#include #include #include #include #include #include #include +#include #include #include #include @@ -38,6 +39,11 @@ namespace CurrentMetrics extern const Metric BrokenDistributedBytesToInsert; } +namespace ProfileEvents +{ + extern const Event DistributedAsyncInsertionFailures; +} + namespace fs = std::filesystem; namespace DB @@ -195,6 +201,15 @@ void DistributedAsyncInsertDirectoryQueue::run() /// No errors while processing existing files. /// Let's see maybe there are more files to process. do_sleep = false; + + const auto now = std::chrono::system_clock::now(); + if (now - last_decrease_time > decrease_error_count_period) + { + std::lock_guard status_lock(status_mutex); + + status.error_count /= 2; + last_decrease_time = now; + } } catch (...) { @@ -213,15 +228,6 @@ void DistributedAsyncInsertDirectoryQueue::run() else LOG_TEST(LogFrequencyLimiter(log, 30), "Skipping send data over distributed table."); - const auto now = std::chrono::system_clock::now(); - if (now - last_decrease_time > decrease_error_count_period) - { - std::lock_guard status_lock(status_mutex); - - status.error_count /= 2; - last_decrease_time = now; - } - if (do_sleep) break; } @@ -376,6 +382,8 @@ try } catch (...) { + ProfileEvents::increment(ProfileEvents::DistributedAsyncInsertionFailures); + std::lock_guard status_lock(status_mutex); ++status.error_count; @@ -516,7 +524,16 @@ void DistributedAsyncInsertDirectoryQueue::processFilesWithBatching() DistributedAsyncInsertBatch batch(*this); batch.deserialize(); - batch.send(); + + /// In case of recovery it is possible that some of files will be + /// missing, if server had been restarted abnormally + /// (between unlink(*.bin) and unlink(current_batch.txt)). + /// + /// But current_batch_file_path should be removed anyway, since if some + /// file was missing, then the batch is not complete and there is no + /// point in trying to pretend that it will not break deduplication. + if (batch.valid()) + batch.send(); auto dir_sync_guard = getDirectorySyncGuard(relative_path); fs::remove(current_batch_file_path); diff --git a/src/Storages/MergeTree/MergeTreeData.cpp b/src/Storages/MergeTree/MergeTreeData.cpp index 2f57ba4bfbc..8445c513372 100644 --- a/src/Storages/MergeTree/MergeTreeData.cpp +++ b/src/Storages/MergeTree/MergeTreeData.cpp @@ -2957,7 +2957,7 @@ void MergeTreeData::checkAlterIsPossible(const AlterCommands & commands, Context if (!settings.allow_non_metadata_alters) { - auto mutation_commands = commands.getMutationCommands(new_metadata, settings.materialize_ttl_after_modify, getContext()); + auto mutation_commands = commands.getMutationCommands(new_metadata, settings.materialize_ttl_after_modify, local_context); if (!mutation_commands.empty()) throw Exception(ErrorCodes::ALTER_OF_COLUMN_IS_FORBIDDEN, @@ -2966,7 +2966,7 @@ void MergeTreeData::checkAlterIsPossible(const AlterCommands & commands, Context queryToString(mutation_commands.ast())); } - commands.apply(new_metadata, getContext()); + commands.apply(new_metadata, local_context); if (commands.hasInvertedIndex(new_metadata) && !settings.allow_experimental_inverted_index) throw Exception(ErrorCodes::SUPPORT_IS_DISABLED, @@ -3253,7 +3253,7 @@ void MergeTreeData::checkAlterIsPossible(const AlterCommands & commands, Context if (!columns_to_check_conversion.empty()) { auto old_header = old_metadata.getSampleBlock(); - performRequiredConversions(old_header, columns_to_check_conversion, getContext()); + performRequiredConversions(old_header, columns_to_check_conversion, local_context); } if (old_metadata.hasSettingsChanges()) @@ -3285,7 +3285,7 @@ void MergeTreeData::checkAlterIsPossible(const AlterCommands & commands, Context } if (setting_name == "storage_policy") - checkStoragePolicy(getContext()->getStoragePolicy(new_value.safeGet())); + checkStoragePolicy(local_context->getStoragePolicy(new_value.safeGet())); } /// Check if it is safe to reset the settings diff --git a/src/Storages/MergeTree/ReplicatedMergeTreeTableMetadata.cpp b/src/Storages/MergeTree/ReplicatedMergeTreeTableMetadata.cpp index 41ff93e28f7..f4e4756279f 100644 --- a/src/Storages/MergeTree/ReplicatedMergeTreeTableMetadata.cpp +++ b/src/Storages/MergeTree/ReplicatedMergeTreeTableMetadata.cpp @@ -409,7 +409,7 @@ StorageInMemoryMetadata ReplicatedMergeTreeTableMetadata::Diff::getNewMetadata(c ParserTTLExpressionList parser; auto ttl_for_table_ast = parseQuery(parser, new_ttl_table, 0, DBMS_DEFAULT_MAX_PARSER_DEPTH); new_metadata.table_ttl = TTLTableDescription::getTTLForTableFromAST( - ttl_for_table_ast, new_metadata.columns, context, new_metadata.primary_key); + ttl_for_table_ast, new_metadata.columns, context, new_metadata.primary_key, true /* allow_suspicious; because it is replication */); } else /// TTL was removed { @@ -422,7 +422,7 @@ StorageInMemoryMetadata ReplicatedMergeTreeTableMetadata::Diff::getNewMetadata(c new_metadata.column_ttls_by_name.clear(); for (const auto & [name, ast] : new_metadata.columns.getColumnTTLs()) { - auto new_ttl_entry = TTLDescription::getTTLFromAST(ast, new_metadata.columns, context, new_metadata.primary_key); + auto new_ttl_entry = TTLDescription::getTTLFromAST(ast, new_metadata.columns, context, new_metadata.primary_key, true /* allow_suspicious; because it is replication */); new_metadata.column_ttls_by_name[name] = new_ttl_entry; } @@ -454,7 +454,7 @@ StorageInMemoryMetadata ReplicatedMergeTreeTableMetadata::Diff::getNewMetadata(c if (!ttl_table_changed && new_metadata.table_ttl.definition_ast != nullptr) new_metadata.table_ttl = TTLTableDescription::getTTLForTableFromAST( - new_metadata.table_ttl.definition_ast, new_metadata.columns, context, new_metadata.primary_key); + new_metadata.table_ttl.definition_ast, new_metadata.columns, context, new_metadata.primary_key, true /* allow_suspicious; because it is replication */); if (!projections_changed) { diff --git a/src/Storages/MergeTree/registerStorageMergeTree.cpp b/src/Storages/MergeTree/registerStorageMergeTree.cpp index 3310b8cb72d..9a5af77d57c 100644 --- a/src/Storages/MergeTree/registerStorageMergeTree.cpp +++ b/src/Storages/MergeTree/registerStorageMergeTree.cpp @@ -581,10 +581,12 @@ static StoragePtr create(const StorageFactory::Arguments & args) if (args.storage_def->sample_by) metadata.sampling_key = KeyDescription::getKeyFromAST(args.storage_def->sample_by->ptr(), metadata.columns, context); + bool allow_suspicious_ttl = args.attach || args.getLocalContext()->getSettingsRef().allow_suspicious_ttl_expressions; + if (args.storage_def->ttl_table) { metadata.table_ttl = TTLTableDescription::getTTLForTableFromAST( - args.storage_def->ttl_table->ptr(), metadata.columns, context, metadata.primary_key); + args.storage_def->ttl_table->ptr(), metadata.columns, context, metadata.primary_key, allow_suspicious_ttl); } if (args.query.columns_list && args.query.columns_list->indices) @@ -602,7 +604,7 @@ static StoragePtr create(const StorageFactory::Arguments & args) auto column_ttl_asts = columns.getColumnTTLs(); for (const auto & [name, ast] : column_ttl_asts) { - auto new_ttl_entry = TTLDescription::getTTLFromAST(ast, columns, context, metadata.primary_key); + auto new_ttl_entry = TTLDescription::getTTLFromAST(ast, columns, context, metadata.primary_key, allow_suspicious_ttl); metadata.column_ttls_by_name[name] = new_ttl_entry; } diff --git a/src/Storages/StorageReplicatedMergeTree.cpp b/src/Storages/StorageReplicatedMergeTree.cpp index ac771cdd0ae..4fb21705534 100644 --- a/src/Storages/StorageReplicatedMergeTree.cpp +++ b/src/Storages/StorageReplicatedMergeTree.cpp @@ -1379,7 +1379,6 @@ void StorageReplicatedMergeTree::setTableStructure(const StorageID & table_id, c checkTTLExpressions(new_metadata, old_metadata); setProperties(new_metadata, old_metadata); - DatabaseCatalog::instance().getDatabase(table_id.database_name)->alterTable(local_context, table_id, new_metadata); } diff --git a/src/Storages/System/StorageSystemServerSettings.cpp b/src/Storages/System/StorageSystemServerSettings.cpp index 3a3acabc5a3..439e3972bc7 100644 --- a/src/Storages/System/StorageSystemServerSettings.cpp +++ b/src/Storages/System/StorageSystemServerSettings.cpp @@ -46,17 +46,20 @@ void StorageSystemServerSettings::fillData(MutableColumns & res_columns, Context {"max_concurrent_insert_queries", std::to_string(context->getProcessList().getMaxInsertQueriesAmount())}, {"max_concurrent_select_queries", std::to_string(context->getProcessList().getMaxSelectQueriesAmount())}, - {"background_pool_size", std::to_string(context->getMergeMutateExecutor()->getMaxThreads())}, - {"background_move_pool_size", std::to_string(context->getMovesExecutor()->getMaxThreads())}, - {"background_fetches_pool_size", std::to_string(context->getFetchesExecutor()->getMaxThreads())}, - {"background_common_pool_size", std::to_string(context->getCommonExecutor()->getMaxThreads())}, - {"background_buffer_flush_schedule_pool_size", std::to_string(CurrentMetrics::get(CurrentMetrics::BackgroundBufferFlushSchedulePoolSize))}, {"background_schedule_pool_size", std::to_string(CurrentMetrics::get(CurrentMetrics::BackgroundSchedulePoolSize))}, {"background_message_broker_schedule_pool_size", std::to_string(CurrentMetrics::get(CurrentMetrics::BackgroundMessageBrokerSchedulePoolSize))}, {"background_distributed_schedule_pool_size", std::to_string(CurrentMetrics::get(CurrentMetrics::BackgroundDistributedSchedulePoolSize))} }; + if (context->areBackgroundExecutorsInitialized()) + { + updated.insert({"background_pool_size", std::to_string(context->getMergeMutateExecutor()->getMaxThreads())}); + updated.insert({"background_move_pool_size", std::to_string(context->getMovesExecutor()->getMaxThreads())}); + updated.insert({"background_fetches_pool_size", std::to_string(context->getFetchesExecutor()->getMaxThreads())}); + updated.insert({"background_common_pool_size", std::to_string(context->getCommonExecutor()->getMaxThreads())}); + } + const auto & config = context->getConfigRef(); ServerSettings settings; settings.loadSettingsFromConfig(config); diff --git a/src/Storages/System/StorageSystemStackTrace.cpp b/src/Storages/System/StorageSystemStackTrace.cpp index 477b784952e..a860930087c 100644 --- a/src/Storages/System/StorageSystemStackTrace.cpp +++ b/src/Storages/System/StorageSystemStackTrace.cpp @@ -57,12 +57,13 @@ std::atomic signal_latch = false; /// Only need for thread sanitizer. /** Notes: * Only one query from the table can be processed at the moment of time. - * This is ensured by the mutex in fillData function. + * This is ensured by the mutex in StorageSystemStackTraceSource. * We obtain information about threads by sending signal and receiving info from the signal handler. * Information is passed via global variables and pipe is used for signaling. * Actually we can send all information via pipe, but we read from it with timeout just in case, - * so it's convenient to use is only for signaling. + * so it's convenient to use it only for signaling. */ +std::mutex mutex; StackTrace stack_trace{NoCapture{}}; @@ -189,7 +190,7 @@ ThreadIdToName getFilteredThreadNames(ASTPtr query, ContextPtr context, const Pa tid_to_name[tid] = thread_name; all_thread_names->insert(thread_name); } - LOG_TEST(log, "Read {} thread names for {} threads, took {} ms", tid_to_name.size(), thread_ids.size(), watch.elapsedMilliseconds()); + LOG_TRACE(log, "Read {} thread names for {} threads, took {} ms", tid_to_name.size(), thread_ids.size(), watch.elapsedMilliseconds()); Block block { ColumnWithTypeAndName(std::move(all_thread_names), std::make_shared(), "thread_name") }; VirtualColumnUtils::filterBlockWithQuery(query, block, context); @@ -229,6 +230,8 @@ public: , pipe_read_timeout_ms(static_cast(context->getSettingsRef().storage_system_stack_trace_pipe_read_timeout_ms.totalMilliseconds())) , log(log_) , proc_it("/proc/self/task") + /// It shouldn't be possible to do concurrent reads from this table. + , lock(mutex) { /// Create a mask of what columns are needed in the result. NameSet names_set(column_names.begin(), column_names.end()); @@ -241,16 +244,13 @@ public: protected: Chunk generate() override { - /// It shouldn't be possible to do concurrent reads from this table. - std::lock_guard lock(mutex); - MutableColumns res_columns = header.cloneEmptyColumns(); ColumnPtr thread_ids; { Stopwatch watch; thread_ids = getFilteredThreadIds(); - LOG_TEST(log, "Read {} threads, took {} ms", thread_ids->size(), watch.elapsedMilliseconds()); + LOG_TRACE(log, "Read {} threads, took {} ms", thread_ids->size(), watch.elapsedMilliseconds()); } if (thread_ids->empty()) return Chunk(); @@ -332,7 +332,7 @@ protected: ++sequence_num; } } - LOG_TEST(log, "Send signal to {} threads (total), took {} ms", signals_sent, signals_sent_ms); + LOG_TRACE(log, "Send signal to {} threads (total), took {} ms", signals_sent, signals_sent_ms); UInt64 num_rows = res_columns.at(0)->size(); Chunk chunk(std::move(res_columns), num_rows); @@ -357,7 +357,7 @@ private: size_t signals_sent = 0; size_t signals_sent_ms = 0; - std::mutex mutex; + std::unique_lock lock; ColumnPtr getFilteredThreadIds() { diff --git a/src/Storages/System/attachSystemTables.cpp b/src/Storages/System/attachSystemTables.cpp index ca49a546b94..2d7e728bbba 100644 --- a/src/Storages/System/attachSystemTables.cpp +++ b/src/Storages/System/attachSystemTables.cpp @@ -108,72 +108,75 @@ namespace DB { +template void attachSystemTablesLocal(ContextPtr context, IDatabase & system_database) { - attach(context, system_database, "one"); - attach(context, system_database, "numbers", false); - attach(context, system_database, "numbers_mt", true); - attach(context, system_database, "zeros", false); - attach(context, system_database, "zeros_mt", true); - attach(context, system_database, "databases"); - attach(context, system_database, "tables"); - attach(context, system_database, "columns"); - attach(context, system_database, "functions"); - attach(context, system_database, "events"); - attach(context, system_database, "settings"); - attach(context, system_database, "server_settings"); - attach(context, system_database, "settings_changes"); - attach>(context, system_database, "merge_tree_settings"); - attach>(context, system_database, "replicated_merge_tree_settings"); - attach(context, system_database, "build_options"); - attach(context, system_database, "formats"); - attach(context, system_database, "table_functions"); - attach(context, system_database, "aggregate_function_combinators"); - attach(context, system_database, "data_type_families"); - attach(context, system_database, "collations"); - attach(context, system_database, "table_engines"); - attach(context, system_database, "contributors"); - attach(context, system_database, "users"); - attach(context, system_database, "roles"); - attach(context, system_database, "grants"); - attach(context, system_database, "role_grants"); - attach(context, system_database, "current_roles"); - attach(context, system_database, "enabled_roles"); - attach(context, system_database, "settings_profiles"); - attach(context, system_database, "settings_profile_elements"); - attach(context, system_database, "row_policies"); - attach(context, system_database, "quotas"); - attach(context, system_database, "quota_limits"); - attach(context, system_database, "quota_usage"); - attach(context, system_database, "quotas_usage"); - attach(context, system_database, "user_directories"); - attach(context, system_database, "privileges"); - attach(context, system_database, "errors"); - attach(context, system_database, "warnings"); - attach(context, system_database, "data_skipping_indices"); - attach(context, system_database, "licenses"); - attach(context, system_database, "time_zones"); - attach(context, system_database, "backups"); - attach(context, system_database, "schema_inference_cache"); - attach(context, system_database, "dropped_tables"); - attach(context, system_database, "scheduler"); + attachLazyOrNot(context, system_database, "one"); + attachLazyOrNot(context, system_database, "numbers", false); + attachLazyOrNot(context, system_database, "numbers_mt", true); + attachLazyOrNot(context, system_database, "zeros", false); + attachLazyOrNot(context, system_database, "zeros_mt", true); + attachLazyOrNot(context, system_database, "databases"); + attachLazyOrNot(context, system_database, "tables"); + attachLazyOrNot(context, system_database, "columns"); + attachLazyOrNot(context, system_database, "functions"); + attachLazyOrNot(context, system_database, "events"); + attachLazyOrNot(context, system_database, "settings"); + attachLazyOrNot(context, system_database, "server_settings"); + attachLazyOrNot(context, system_database, "settings_changes"); + attachLazyOrNot>(context, system_database, "merge_tree_settings"); + attachLazyOrNot>(context, system_database, "replicated_merge_tree_settings"); + attachLazyOrNot(context, system_database, "build_options"); + attachLazyOrNot(context, system_database, "formats"); + attachLazyOrNot(context, system_database, "table_functions"); + attachLazyOrNot(context, system_database, "aggregate_function_combinators"); + attachLazyOrNot(context, system_database, "data_type_families"); + attachLazyOrNot(context, system_database, "collations"); + attachLazyOrNot(context, system_database, "table_engines"); + attachLazyOrNot(context, system_database, "contributors"); + attachLazyOrNot(context, system_database, "users"); + attachLazyOrNot(context, system_database, "roles"); + attachLazyOrNot(context, system_database, "grants"); + attachLazyOrNot(context, system_database, "role_grants"); + attachLazyOrNot(context, system_database, "current_roles"); + attachLazyOrNot(context, system_database, "enabled_roles"); + attachLazyOrNot(context, system_database, "settings_profiles"); + attachLazyOrNot(context, system_database, "settings_profile_elements"); + attachLazyOrNot(context, system_database, "row_policies"); + attachLazyOrNot(context, system_database, "quotas"); + attachLazyOrNot(context, system_database, "quota_limits"); + attachLazyOrNot(context, system_database, "quota_usage"); + attachLazyOrNot(context, system_database, "quotas_usage"); + attachLazyOrNot(context, system_database, "user_directories"); + attachLazyOrNot(context, system_database, "privileges"); + attachLazyOrNot(context, system_database, "errors"); + attachLazyOrNot(context, system_database, "warnings"); + attachLazyOrNot(context, system_database, "data_skipping_indices"); + attachLazyOrNot(context, system_database, "licenses"); + attachLazyOrNot(context, system_database, "time_zones"); + attachLazyOrNot(context, system_database, "backups"); + attachLazyOrNot(context, system_database, "schema_inference_cache"); + attachLazyOrNot(context, system_database, "dropped_tables"); + attachLazyOrNot(context, system_database, "scheduler"); #if defined(__ELF__) && !defined(OS_FREEBSD) - attach(context, system_database, "symbols"); + attachLazyOrNot(context, system_database, "symbols"); #endif #if USE_RDKAFKA - attach(context, system_database, "kafka_consumers"); + attachLazyOrNot(context, system_database, "kafka_consumers"); #endif #ifdef OS_LINUX - attach(context, system_database, "stack_trace"); + attachLazyOrNot(context, system_database, "stack_trace"); #endif #if USE_ROCKSDB - attach(context, system_database, "rocksdb"); + attachLazyOrNot(context, system_database, "rocksdb"); #endif } +template void attachSystemTablesLocal(ContextPtr context, IDatabase & system_database); +template void attachSystemTablesLocal(ContextPtr context, IDatabase & system_database); void attachSystemTablesServer(ContextPtr context, IDatabase & system_database, bool has_zookeeper) { - attachSystemTablesLocal(context, system_database); + attachSystemTablesLocal(context, system_database); attach(context, system_database, "parts"); attach(context, system_database, "projection_parts"); diff --git a/src/Storages/System/attachSystemTables.h b/src/Storages/System/attachSystemTables.h index 4c1a79f84dd..d09189ca92f 100644 --- a/src/Storages/System/attachSystemTables.h +++ b/src/Storages/System/attachSystemTables.h @@ -9,8 +9,12 @@ namespace DB class AsynchronousMetrics; class IDatabase; -void attachSystemTablesServer(ContextPtr context, IDatabase & system_database, bool has_zookeeper); +template void attachSystemTablesLocal(ContextPtr context, IDatabase & system_database); +void attachSystemTablesServer(ContextPtr context, IDatabase & system_database, bool has_zookeeper); void attachSystemTablesAsync(ContextPtr context, IDatabase & system_database, AsynchronousMetrics & async_metrics); +extern template void attachSystemTablesLocal(ContextPtr context, IDatabase & system_database); +extern template void attachSystemTablesLocal(ContextPtr context, IDatabase & system_database); + } diff --git a/src/Storages/System/attachSystemTablesImpl.h b/src/Storages/System/attachSystemTablesImpl.h index a1fae985d92..45fd7957513 100644 --- a/src/Storages/System/attachSystemTablesImpl.h +++ b/src/Storages/System/attachSystemTablesImpl.h @@ -1,15 +1,42 @@ #pragma once #include +#include #include namespace DB { +template +void attachLazy(ContextPtr context, IDatabase & system_database, const String & table_name, StorageArgs && ... args) +{ + if (system_database.getUUID() == UUIDHelpers::Nil) + { + /// Attach to Ordinary database. + auto table_id = StorageID(DatabaseCatalog::SYSTEM_DATABASE, table_name); + system_database.registerLazyTable(context, table_name, [table_id, ... captured_args = std::forward(args)] mutable + { + return std::make_shared(table_id, std::forward(captured_args)...); + }); + } + else + { + /// Attach to Atomic database. + /// NOTE: UUIDs are not persistent, but it's ok since no data are stored on disk for these storages + /// and path is actually not used + auto table_id = StorageID(DatabaseCatalog::SYSTEM_DATABASE, table_name, UUIDHelpers::generateV4()); + DatabaseCatalog::instance().addUUIDMapping(table_id.uuid); + String path = "store/" + DatabaseCatalog::getPathForUUID(table_id.uuid); + system_database.registerLazyTable(context, table_name, [table_id, ... captured_args = std::forward(args)] mutable + { + return std::make_shared(table_id, std::forward(captured_args)...); + }, path); + } +} + template void attach(ContextPtr context, IDatabase & system_database, const String & table_name, StorageArgs && ... args) { - assert(system_database.getDatabaseName() == DatabaseCatalog::SYSTEM_DATABASE); if (system_database.getUUID() == UUIDHelpers::Nil) { /// Attach to Ordinary database. @@ -28,4 +55,15 @@ void attach(ContextPtr context, IDatabase & system_database, const String & tabl } } +template +void attachLazyOrNot(ContextPtr context, IDatabase & system_database, const String & table_name, StorageArgs && ... args) +{ + assert(system_database.getDatabaseName() == DatabaseCatalog::SYSTEM_DATABASE); + + if constexpr (lazy) + attachLazy(context, system_database, table_name, std::forward(args)...); + else + attach(context, system_database, table_name, std::forward(args)...); +} + } diff --git a/src/Storages/TTLDescription.cpp b/src/Storages/TTLDescription.cpp index f601fed06ac..41a222525bf 100644 --- a/src/Storages/TTLDescription.cpp +++ b/src/Storages/TTLDescription.cpp @@ -55,22 +55,29 @@ TTLAggregateDescription & TTLAggregateDescription::operator=(const TTLAggregateD namespace { -void checkTTLExpression(const ExpressionActionsPtr & ttl_expression, const String & result_column_name) +void checkTTLExpression(const ExpressionActionsPtr & ttl_expression, const String & result_column_name, bool allow_suspicious) { - for (const auto & action : ttl_expression->getActions()) + /// Do not apply this check in ATTACH queries for compatibility reasons and if explicitly allowed. + if (!allow_suspicious) { - if (action.node->type == ActionsDAG::ActionType::FUNCTION) + if (ttl_expression->getRequiredColumns().empty()) + throw Exception(ErrorCodes::BAD_ARGUMENTS, + "TTL expression {} does not depend on any of the columns of the table", result_column_name); + + for (const auto & action : ttl_expression->getActions()) { - const IFunctionBase & func = *action.node->function_base; - if (!func.isDeterministic()) - throw Exception(ErrorCodes::BAD_ARGUMENTS, - "TTL expression cannot contain non-deterministic functions, but contains function {}", - func.getName()); + if (action.node->type == ActionsDAG::ActionType::FUNCTION) + { + const IFunctionBase & func = *action.node->function_base; + if (!func.isDeterministic()) + throw Exception(ErrorCodes::BAD_ARGUMENTS, + "TTL expression cannot contain non-deterministic functions, but contains function {}", + func.getName()); + } } } const auto & result_column = ttl_expression->getSampleBlock().getByName(result_column_name); - if (!typeid_cast(result_column.type.get()) && !typeid_cast(result_column.type.get())) { @@ -162,7 +169,8 @@ TTLDescription TTLDescription::getTTLFromAST( const ASTPtr & definition_ast, const ColumnsDescription & columns, ContextPtr context, - const KeyDescription & primary_key) + const KeyDescription & primary_key, + bool is_attach) { TTLDescription result; const auto * ttl_element = definition_ast->as(); @@ -289,7 +297,7 @@ TTLDescription TTLDescription::getTTLFromAST( } } - checkTTLExpression(result.expression, result.result_column); + checkTTLExpression(result.expression, result.result_column, is_attach || context->getSettingsRef().allow_suspicious_ttl_expressions); return result; } @@ -327,7 +335,8 @@ TTLTableDescription TTLTableDescription::getTTLForTableFromAST( const ASTPtr & definition_ast, const ColumnsDescription & columns, ContextPtr context, - const KeyDescription & primary_key) + const KeyDescription & primary_key, + bool is_attach) { TTLTableDescription result; if (!definition_ast) @@ -338,7 +347,7 @@ TTLTableDescription TTLTableDescription::getTTLForTableFromAST( bool have_unconditional_delete_ttl = false; for (const auto & ttl_element_ptr : definition_ast->children) { - auto ttl = TTLDescription::getTTLFromAST(ttl_element_ptr, columns, context, primary_key); + auto ttl = TTLDescription::getTTLFromAST(ttl_element_ptr, columns, context, primary_key, is_attach); if (ttl.mode == TTLMode::DELETE) { if (!ttl.where_expression) @@ -380,7 +389,7 @@ TTLTableDescription TTLTableDescription::parse(const String & str, const Columns ASTPtr ast = parseQuery(parser, str, 0, DBMS_DEFAULT_MAX_PARSER_DEPTH); FunctionNameNormalizer().visit(ast.get()); - return getTTLForTableFromAST(ast, columns, context, primary_key); + return getTTLForTableFromAST(ast, columns, context, primary_key, context->getSettingsRef().allow_suspicious_ttl_expressions); } } diff --git a/src/Storages/TTLDescription.h b/src/Storages/TTLDescription.h index 8f60eb604b5..aab5b43e53e 100644 --- a/src/Storages/TTLDescription.h +++ b/src/Storages/TTLDescription.h @@ -1,4 +1,5 @@ #pragma once + #include #include #include @@ -7,6 +8,7 @@ #include #include + namespace DB { @@ -82,9 +84,8 @@ struct TTLDescription /// Codec name which will be used to recompress data ASTPtr recompression_codec; - /// Parse TTL structure from definition. Able to parse both column and table - /// TTLs. - static TTLDescription getTTLFromAST(const ASTPtr & definition_ast, const ColumnsDescription & columns, ContextPtr context, const KeyDescription & primary_key); + /// Parse TTL structure from definition. Able to parse both column and table TTLs. + static TTLDescription getTTLFromAST(const ASTPtr & definition_ast, const ColumnsDescription & columns, ContextPtr context, const KeyDescription & primary_key, bool is_attach); TTLDescription() = default; TTLDescription(const TTLDescription & other); @@ -121,7 +122,7 @@ struct TTLTableDescription TTLTableDescription & operator=(const TTLTableDescription & other); static TTLTableDescription getTTLForTableFromAST( - const ASTPtr & definition_ast, const ColumnsDescription & columns, ContextPtr context, const KeyDescription & primary_key); + const ASTPtr & definition_ast, const ColumnsDescription & columns, ContextPtr context, const KeyDescription & primary_key, bool is_attach); /// Parse description from string static TTLTableDescription parse(const String & str, const ColumnsDescription & columns, ContextPtr context, const KeyDescription & primary_key); diff --git a/tests/ci/ci_config.py b/tests/ci/ci_config.py index d1d2021753e..aa3aa5654aa 100644 --- a/tests/ci/ci_config.py +++ b/tests/ci/ci_config.py @@ -208,6 +208,13 @@ CI_CONFIG = CiConfig( static_binary_name="amd64compat", comment="SSE2-only build", ), + "binary_amd64_musl": BuildConfig( + name="binary_amd64_musl", + compiler="clang-17-amd64-musl", + package_type="binary", + static_binary_name="amd64musl", + comment="Build with Musl", + ), "binary_riscv64": BuildConfig( name="binary_riscv64", compiler="clang-17-riscv64", @@ -249,6 +256,7 @@ CI_CONFIG = CiConfig( "binary_riscv64", "binary_s390x", "binary_amd64_compat", + "binary_amd64_musl", ], }, test_configs={ diff --git a/tests/clickhouse-test b/tests/clickhouse-test index 115e5ac7ba3..9599a1d1723 100755 --- a/tests/clickhouse-test +++ b/tests/clickhouse-test @@ -2205,7 +2205,7 @@ def reportLogStats(args): GROUP BY message_format_string ORDER BY count DESC LIMIT 100 - FORMAT TSVWithNamesAndTypes + FORMAT PrettySpaceNoEscapes """ value = clickhouse_execute(args, query).decode(errors="replace") print("\nTop patterns of log messages:\n") @@ -2225,7 +2225,7 @@ def reportLogStats(args): GROUP BY pattern ORDER BY count DESC LIMIT 30 - FORMAT TSVWithNamesAndTypes + FORMAT PrettySpaceNoEscapes """ value = clickhouse_execute(args, query).decode(errors="replace") print("\nTop messages without format string (fmt::runtime):\n") @@ -2238,10 +2238,10 @@ def reportLogStats(args): WHERE (now() - toIntervalMinute(240)) < event_time AND (message NOT LIKE (replaceRegexpAll(message_format_string, '{[:.0-9dfx]*}', '%') AS s)) AND (message NOT LIKE concat('%Exception: ', s, '%')) - GROUP BY message_format_string ORDER BY count() DESC LIMIT 20 FORMAT TSVWithNamesAndTypes + GROUP BY message_format_string ORDER BY count() DESC LIMIT 20 FORMAT PrettySpaceNoEscapes """ value = clickhouse_execute(args, query).decode(errors="replace") - print("\nTop messages that does not match its format string:\n") + print("\nTop messages not matching their format strings:\n") print(value) print("\n") @@ -2269,13 +2269,13 @@ def reportLogStats(args): 'Attempt to read after eof', 'String size is too big ({}), maximum: {}' ) AS known_short_messages SELECT count() AS c, message_format_string, substr(any(message), 1, 120), - min(if(length(regexpExtract(message, '(.*)\\([A-Z0-9_]+\\)')) as pref > 0, pref, length(message)) - 26 AS length_without_exception_boilerplate) AS min_length_without_exception_boilerplate + min(if(notEmpty(regexpExtract(message, '(.*)\\([A-Z0-9_]+\\)') as prefix), prefix, length(message)) - 26 AS length_without_exception_boilerplate) AS min_length_without_exception_boilerplate FROM system.text_log WHERE (now() - toIntervalMinute(240)) < event_time AND (length(message_format_string) < 16 - OR (message ilike '%DB::Exception%' AND length_without_exception_boilerplate < 30)) + OR (message ILIKE '%DB::Exception%' AND length_without_exception_boilerplate < 30)) AND message_format_string NOT IN known_short_messages - GROUP BY message_format_string ORDER BY c DESC LIMIT 50 FORMAT TSVWithNamesAndTypes + GROUP BY message_format_string ORDER BY c DESC LIMIT 50 FORMAT PrettySpaceNoEscapes """ value = clickhouse_execute(args, query).decode(errors="replace") print("\nTop short messages:\n") diff --git a/tests/integration/helpers/cluster.py b/tests/integration/helpers/cluster.py index 5e4bb32cf94..c0b145b047a 100644 --- a/tests/integration/helpers/cluster.py +++ b/tests/integration/helpers/cluster.py @@ -390,6 +390,7 @@ class ClickHouseCluster: odbc_bridge_bin_path=None, library_bridge_bin_path=None, zookeeper_config_path=None, + keeper_config_dir=None, custom_dockerd_host=None, zookeeper_keyfile=None, zookeeper_certfile=None, @@ -426,6 +427,12 @@ class ClickHouseCluster: else p.join(HELPERS_DIR, "zookeeper_config.xml") ) + self.keeper_config_dir = ( + p.join(self.base_dir, keeper_config_dir) + if keeper_config_dir + else HELPERS_DIR + ) + project_name = ( pwd.getpwuid(os.getuid()).pw_name + p.basename(self.base_dir) + self.name ) @@ -2725,7 +2732,9 @@ class ClickHouseCluster: if self.use_keeper: # TODO: remove hardcoded paths from here for i in range(1, 4): shutil.copy( - os.path.join(HELPERS_DIR, f"keeper_config{i}.xml"), + os.path.join( + self.keeper_config_dir, f"keeper_config{i}.xml" + ), os.path.join( self.keeper_instance_dir_prefix + f"{i}", "config" ), diff --git a/tests/integration/helpers/postgres_utility.py b/tests/integration/helpers/postgres_utility.py index 4ea4cefe5a4..4bf549174e8 100644 --- a/tests/integration/helpers/postgres_utility.py +++ b/tests/integration/helpers/postgres_utility.py @@ -193,11 +193,11 @@ class PostgresManager: database_name = self.database_or_default(database_name) self.drop_postgres_db(database_name) self.created_postgres_db_list.add(database_name) - self.cursor.execute(f"CREATE DATABASE {database_name}") + self.cursor.execute(f'CREATE DATABASE "{database_name}"') def drop_postgres_db(self, database_name=""): database_name = self.database_or_default(database_name) - self.cursor.execute(f"DROP DATABASE IF EXISTS {database_name} WITH (FORCE)") + self.cursor.execute(f'DROP DATABASE IF EXISTS "{database_name}" WITH (FORCE)') if database_name in self.created_postgres_db_list: self.created_postgres_db_list.remove(database_name) @@ -216,19 +216,19 @@ class PostgresManager: if len(schema_name) == 0: self.instance.query( f""" - CREATE DATABASE {database_name} + CREATE DATABASE \"{database_name}\" ENGINE = PostgreSQL('{self.ip}:{self.port}', '{postgres_database}', 'postgres', 'mysecretpassword')""" ) else: self.instance.query( f""" - CREATE DATABASE {database_name} + CREATE DATABASE \"{database_name}\" ENGINE = PostgreSQL('{self.ip}:{self.port}', '{postgres_database}', 'postgres', 'mysecretpassword', '{schema_name}')""" ) def drop_clickhouse_postgres_db(self, database_name=""): database_name = self.database_or_default(database_name) - self.instance.query(f"DROP DATABASE IF EXISTS {database_name}") + self.instance.query(f'DROP DATABASE IF EXISTS "{database_name}"') if database_name in self.created_ch_postgres_db_list: self.created_ch_postgres_db_list.remove(database_name) @@ -368,7 +368,7 @@ def check_tables_are_synchronized( result_query = f"select * from {table_path} order by {order_by};" expected = instance.query( - f"select * from {postgres_database}.{table_name} order by {order_by};" + f"select * from `{postgres_database}`.`{table_name}` order by {order_by};" ) result = instance.query(result_query) @@ -382,7 +382,7 @@ def check_tables_are_synchronized( if result != expected: count = int(instance.query(f"select count() from {table_path}")) expected_count = int( - instance.query(f"select count() from {postgres_database}.{table_name}") + instance.query(f"select count() from `{postgres_database}`.`{table_name}`") ) print(f"Having {count}, expected {expected_count}") assert result == expected diff --git a/tests/integration/test_dictionaries_update_and_reload/test.py b/tests/integration/test_dictionaries_update_and_reload/test.py index 99d08f1b6ea..3d96d0b8dd4 100644 --- a/tests/integration/test_dictionaries_update_and_reload/test.py +++ b/tests/integration/test_dictionaries_update_and_reload/test.py @@ -92,16 +92,16 @@ def test_reload_while_loading(started_cluster): assert get_status("slow") == "NOT_LOADED" assert get_loading_duration("slow") == 0 - # It's not possible to get a value from the dictionary within 0.5 second, so the following query fails by timeout. + # It's not possible to get a value from the dictionary within 1 second, so the following query fails by timeout. with pytest.raises(QueryTimeoutExceedException): - query("SELECT dictGetInt32('slow', 'a', toUInt64(5))", timeout=0.5) + query("SELECT dictGetInt32('slow', 'a', toUInt64(5))", timeout=1) # The dictionary is now loading. assert get_status("slow") == "LOADING" start_time, duration = get_loading_start_time("slow"), get_loading_duration("slow") assert duration > 0 - time.sleep(0.5) # Still loading. + time.sleep(1) # Still loading. assert get_status("slow") == "LOADING" prev_start_time, prev_duration = start_time, duration start_time, duration = get_loading_start_time("slow"), get_loading_duration("slow") @@ -110,14 +110,14 @@ def test_reload_while_loading(started_cluster): # SYSTEM RELOAD DICTIONARY should restart loading. with pytest.raises(QueryTimeoutExceedException): - query("SYSTEM RELOAD DICTIONARY 'slow'", timeout=0.5) + query("SYSTEM RELOAD DICTIONARY 'slow'", timeout=1) assert get_status("slow") == "LOADING" prev_start_time, prev_duration = start_time, duration start_time, duration = get_loading_start_time("slow"), get_loading_duration("slow") assert start_time > prev_start_time assert duration < prev_duration - time.sleep(0.5) # Still loading. + time.sleep(1) # Still loading. assert get_status("slow") == "LOADING" prev_start_time, prev_duration = start_time, duration start_time, duration = get_loading_start_time("slow"), get_loading_duration("slow") @@ -128,7 +128,7 @@ def test_reload_while_loading(started_cluster): replace_in_file_in_container( "/etc/clickhouse-server/dictionaries/slow.xml", "sleep 100", "sleep 0" ) - time.sleep(5) # Configuration files are reloaded once in 5 seconds. + query("SYSTEM RELOAD CONFIG") # This time loading should finish quickly. assert get_status("slow") == "LOADED" diff --git a/tests/queries/0_stateless/02896_memory_accounting_for_user.reference b/tests/integration/test_keeper_memory_soft_limit/__init__.py similarity index 100% rename from tests/queries/0_stateless/02896_memory_accounting_for_user.reference rename to tests/integration/test_keeper_memory_soft_limit/__init__.py diff --git a/tests/integration/test_keeper_memory_soft_limit/configs/keeper_config1.xml b/tests/integration/test_keeper_memory_soft_limit/configs/keeper_config1.xml new file mode 100644 index 00000000000..fe45d09d915 --- /dev/null +++ b/tests/integration/test_keeper_memory_soft_limit/configs/keeper_config1.xml @@ -0,0 +1,49 @@ + + true + :: + 0.0.0.0 + + + trace + /var/log/clickhouse-keeper/clickhouse-keeper.log + /var/log/clickhouse-keeper/clickhouse-keeper.err.log + + + + 2181 + + az-zoo1 + + 1 + + + 10000 + 15000 + trace + false + 2000 + 4000 + 200000000 + + 1 + + + + + 1 + zoo1 + 9444 + + + 2 + zoo2 + 9444 + + + 3 + zoo3 + 9444 + + + + diff --git a/tests/integration/test_keeper_memory_soft_limit/configs/keeper_config2.xml b/tests/integration/test_keeper_memory_soft_limit/configs/keeper_config2.xml new file mode 100644 index 00000000000..f7f6a5718b5 --- /dev/null +++ b/tests/integration/test_keeper_memory_soft_limit/configs/keeper_config2.xml @@ -0,0 +1,50 @@ + + true + :: + 0.0.0.0 + + + trace + /var/log/clickhouse-keeper/clickhouse-keeper.log + /var/log/clickhouse-keeper/clickhouse-keeper.err.log + + + + 2181 + 2 + + az-zoo2 + 1 + + + + 10000 + 15000 + trace + false + 2000 + 4000 + 20000000 + + 1 + + + + + 1 + zoo1 + 9444 + + + 2 + zoo2 + 9444 + + + 3 + zoo3 + 9444 + + + + diff --git a/tests/integration/test_keeper_memory_soft_limit/configs/keeper_config3.xml b/tests/integration/test_keeper_memory_soft_limit/configs/keeper_config3.xml new file mode 100644 index 00000000000..82345aebc46 --- /dev/null +++ b/tests/integration/test_keeper_memory_soft_limit/configs/keeper_config3.xml @@ -0,0 +1,46 @@ + + true + :: + 0.0.0.0 + + + trace + /var/log/clickhouse-keeper/clickhouse-keeper.log + /var/log/clickhouse-keeper/clickhouse-keeper.err.log + + + + 2181 + 3 + + + 10000 + 15000 + trace + false + 2000 + 4000 + 20000000 + + 1 + + + + + 1 + zoo1 + 9444 + + + 2 + zoo2 + 9444 + + + 3 + zoo3 + 9444 + + + + diff --git a/tests/integration/test_keeper_memory_soft_limit/test.py b/tests/integration/test_keeper_memory_soft_limit/test.py new file mode 100644 index 00000000000..d6f3d013a7b --- /dev/null +++ b/tests/integration/test_keeper_memory_soft_limit/test.py @@ -0,0 +1,63 @@ +#!/usr/bin/env python3 +import random +import string +import pytest +from helpers.cluster import ClickHouseCluster +from helpers import keeper_utils +from kazoo.client import KazooClient, KazooState +from kazoo.exceptions import ConnectionLoss + +cluster = ClickHouseCluster(__file__, keeper_config_dir="configs/") + +# clickhouse itself will use external zookeeper +node = cluster.add_instance( + "node", + stay_alive=True, + with_zookeeper=True, +) + + +def random_string(length): + return "".join(random.choices(string.ascii_lowercase + string.digits, k=length)) + + +def get_connection_zk(nodename, timeout=30.0): + _fake_zk_instance = KazooClient( + hosts=cluster.get_instance_ip(nodename) + ":2181", timeout=timeout + ) + _fake_zk_instance.start() + return _fake_zk_instance + + +@pytest.fixture(scope="module") +def started_cluster(): + try: + cluster.start() + + yield cluster + + finally: + cluster.shutdown() + + +def test_soft_limit_create(started_cluster): + started_cluster.wait_zookeeper_to_start() + try: + node_zk = get_connection_zk("zoo1") + loop_time = 100000 + node_zk.create("/test_soft_limit", b"abc") + + for i in range(loop_time): + node_zk.create( + "/test_soft_limit/node_" + str(i), random_string(1000).encode() + ) + except ConnectionLoss: + txn = node_zk.transaction() + for i in range(10): + txn.delete("/test_soft_limit/node_" + str(i)) + + txn.create("/test_soft_limit/node_1000001" + str(i), b"abcde") + txn.commit() + return + + raise Exception("all records are inserted but no error occurs") diff --git a/tests/integration/test_multiple_disks/test.py b/tests/integration/test_multiple_disks/test.py index 30669feb6b3..fdd81284b2a 100644 --- a/tests/integration/test_multiple_disks/test.py +++ b/tests/integration/test_multiple_disks/test.py @@ -1837,7 +1837,8 @@ def _insert_merge_execute( SETTINGS storage_policy='{policy}' """.format( name=name, policy=policy - ) + ), + settings={"allow_suspicious_ttl_expressions": 1}, ) for i in range(parts): diff --git a/tests/integration/test_postgresql_replica_database_engine_2/test.py b/tests/integration/test_postgresql_replica_database_engine_2/test.py index ddfb5608336..e555ae388d9 100644 --- a/tests/integration/test_postgresql_replica_database_engine_2/test.py +++ b/tests/integration/test_postgresql_replica_database_engine_2/test.py @@ -59,6 +59,7 @@ instance2 = cluster.add_instance( pg_manager = PostgresManager() pg_manager2 = PostgresManager() pg_manager_instance2 = PostgresManager() +pg_manager3 = PostgresManager() @pytest.fixture(scope="module") @@ -81,6 +82,12 @@ def started_cluster(): pg_manager2.init( instance2, cluster.postgres_ip, cluster.postgres_port, "postgres_database2" ) + pg_manager3.init( + instance, + cluster.postgres_ip, + cluster.postgres_port, + default_database="postgres-postgres", + ) yield cluster @@ -915,6 +922,28 @@ def test_failed_load_from_snapshot(started_cluster): ) +def test_symbols_in_publication_name(started_cluster): + table = "test_symbols_in_publication_name" + + pg_manager3.create_postgres_table(table) + instance.query( + f"INSERT INTO `{pg_manager3.get_default_database()}`.`{table}` SELECT number, number from numbers(0, 50)" + ) + + pg_manager3.create_materialized_db( + ip=started_cluster.postgres_ip, + port=started_cluster.postgres_port, + settings=[ + f"materialized_postgresql_tables_list = '{table}'", + "materialized_postgresql_backoff_min_ms = 100", + "materialized_postgresql_backoff_max_ms = 100", + ], + ) + check_tables_are_synchronized( + instance, table, postgres_database=pg_manager3.get_default_database() + ) + + def test_generated_columns(started_cluster): table = "test_generated_columns" diff --git a/tests/integration/test_ttl_move/test.py b/tests/integration/test_ttl_move/test.py index 2f18a0a1afa..1df9bc6d3c9 100644 --- a/tests/integration/test_ttl_move/test.py +++ b/tests/integration/test_ttl_move/test.py @@ -302,7 +302,8 @@ def test_moves_work_after_storage_policy_change(started_cluster, name, engine): node1.query( """ALTER TABLE {name} MODIFY TTL now()-3600 TO DISK 'jbod1', d1 TO DISK 'external'""".format( name=name - ) + ), + settings={"allow_suspicious_ttl_expressions": 1}, ) wait_expire_1 = 12 diff --git a/tests/queries/0_stateless/00933_ttl_simple.reference b/tests/queries/0_stateless/00933_ttl_simple.reference index e3982814eab..72f5134e235 100644 --- a/tests/queries/0_stateless/00933_ttl_simple.reference +++ b/tests/queries/0_stateless/00933_ttl_simple.reference @@ -6,11 +6,11 @@ 2000-10-10 00:00:00 0 2100-10-10 00:00:00 3 2100-10-10 2 -CREATE TABLE default.ttl_00933_1\n(\n `b` Int32,\n `a` Int32 TTL now() - 1000\n)\nENGINE = MergeTree\nPARTITION BY tuple()\nORDER BY tuple()\nSETTINGS min_bytes_for_wide_part = 0, index_granularity = 8192 +CREATE TABLE default.ttl_00933_1\n(\n `b` Int32,\n `a` Int32 TTL CAST(\'2000-10-10 00:00:00\', \'DateTime\')\n)\nENGINE = MergeTree\nPARTITION BY tuple()\nORDER BY tuple()\nSETTINGS min_bytes_for_wide_part = 0, index_granularity = 8192 1 0 -CREATE TABLE default.ttl_00933_1\n(\n `b` Int32,\n `a` Int32 TTL now() + 1000\n)\nENGINE = MergeTree\nPARTITION BY tuple()\nORDER BY tuple()\nSETTINGS min_bytes_for_wide_part = 0, index_granularity = 8192 +CREATE TABLE default.ttl_00933_1\n(\n `b` Int32,\n `a` Int32 TTL CAST(\'2100-10-10 00:00:00\', \'DateTime\')\n)\nENGINE = MergeTree\nPARTITION BY tuple()\nORDER BY tuple()\nSETTINGS min_bytes_for_wide_part = 0, index_granularity = 8192 1 1 -CREATE TABLE default.ttl_00933_1\n(\n `b` Int32,\n `a` Int32 TTL today() - 1\n)\nENGINE = MergeTree\nPARTITION BY tuple()\nORDER BY tuple()\nSETTINGS min_bytes_for_wide_part = 0, index_granularity = 8192 +CREATE TABLE default.ttl_00933_1\n(\n `b` Int32,\n `a` Int32 TTL CAST(\'2000-10-10\', \'Date\')\n)\nENGINE = MergeTree\nPARTITION BY tuple()\nORDER BY tuple()\nSETTINGS min_bytes_for_wide_part = 0, index_granularity = 8192 1 0 -CREATE TABLE default.ttl_00933_1\n(\n `b` Int32,\n `a` Int32 TTL today() + 1\n)\nENGINE = MergeTree\nPARTITION BY tuple()\nORDER BY tuple()\nSETTINGS min_bytes_for_wide_part = 0, index_granularity = 8192 +CREATE TABLE default.ttl_00933_1\n(\n `b` Int32,\n `a` Int32 TTL CAST(\'2100-10-10\', \'Date\')\n)\nENGINE = MergeTree\nPARTITION BY tuple()\nORDER BY tuple()\nSETTINGS min_bytes_for_wide_part = 0, index_granularity = 8192 1 1 diff --git a/tests/queries/0_stateless/00933_ttl_simple.sql b/tests/queries/0_stateless/00933_ttl_simple.sql index ad40e7c7e47..c1df338a0ff 100644 --- a/tests/queries/0_stateless/00933_ttl_simple.sql +++ b/tests/queries/0_stateless/00933_ttl_simple.sql @@ -8,7 +8,8 @@ -- ┌───────────────now()─┬─toDate(toTimeZone(now(), 'America/Mazatlan'))─┬────today()─┐ -- │ 2023-07-24 06:24:06 │ 2023-07-23 │ 2023-07-24 │ -- └─────────────────────┴───────────────────────────────────────────────┴────────────┘ -set session_timezone = ''; +SET session_timezone = ''; +SET allow_suspicious_ttl_expressions = 1; drop table if exists ttl_00933_1; @@ -65,7 +66,9 @@ select * from ttl_00933_1 order by d; -- const DateTime TTL positive drop table if exists ttl_00933_1; -create table ttl_00933_1 (b Int, a Int ttl now()-1000) engine = MergeTree order by tuple() partition by tuple() settings min_bytes_for_wide_part = 0; +create table ttl_00933_1 (b Int, a Int ttl '2000-10-10 00:00:00'::DateTime) +engine = MergeTree order by tuple() partition by tuple() settings min_bytes_for_wide_part = 0; + show create table ttl_00933_1; insert into ttl_00933_1 values (1, 1); optimize table ttl_00933_1 final; @@ -73,7 +76,7 @@ select * from ttl_00933_1; -- const DateTime TTL negative drop table if exists ttl_00933_1; -create table ttl_00933_1 (b Int, a Int ttl now()+1000) engine = MergeTree order by tuple() partition by tuple() settings min_bytes_for_wide_part = 0; +create table ttl_00933_1 (b Int, a Int ttl '2100-10-10 00:00:00'::DateTime) engine = MergeTree order by tuple() partition by tuple() settings min_bytes_for_wide_part = 0; show create table ttl_00933_1; insert into ttl_00933_1 values (1, 1); optimize table ttl_00933_1 final; @@ -81,7 +84,7 @@ select * from ttl_00933_1; -- const Date TTL positive drop table if exists ttl_00933_1; -create table ttl_00933_1 (b Int, a Int ttl today()-1) engine = MergeTree order by tuple() partition by tuple() settings min_bytes_for_wide_part = 0; +create table ttl_00933_1 (b Int, a Int ttl '2000-10-10'::Date) engine = MergeTree order by tuple() partition by tuple() settings min_bytes_for_wide_part = 0; show create table ttl_00933_1; insert into ttl_00933_1 values (1, 1); optimize table ttl_00933_1 final; @@ -89,7 +92,7 @@ select * from ttl_00933_1; -- const Date TTL negative drop table if exists ttl_00933_1; -create table ttl_00933_1 (b Int, a Int ttl today()+1) engine = MergeTree order by tuple() partition by tuple() settings min_bytes_for_wide_part = 0; +create table ttl_00933_1 (b Int, a Int ttl '2100-10-10'::Date) engine = MergeTree order by tuple() partition by tuple() settings min_bytes_for_wide_part = 0; show create table ttl_00933_1; insert into ttl_00933_1 values (1, 1); optimize table ttl_00933_1 final; diff --git a/tests/queries/0_stateless/01051_system_stack_trace.reference b/tests/queries/0_stateless/01051_system_stack_trace.reference deleted file mode 100644 index 29f6990e6b4..00000000000 --- a/tests/queries/0_stateless/01051_system_stack_trace.reference +++ /dev/null @@ -1,18 +0,0 @@ --- { echo } -SELECT count() > 0 FROM system.stack_trace WHERE query_id != '' AND thread_name = 'TCPHandler'; -1 --- opimization for not reading /proc/self/task/{}/comm and avoid sending signal -SELECT countIf(thread_id > 0) > 0 FROM system.stack_trace; -1 --- optimization for trace -SELECT length(trace) > 0 FROM system.stack_trace WHERE length(trace) > 0 LIMIT 1; -1 --- optimization for query_id -SELECT length(query_id) > 0 FROM system.stack_trace WHERE query_id != '' AND thread_name = 'TCPHandler' LIMIT 1; -1 --- optimization for thread_name -SELECT length(thread_name) > 0 FROM system.stack_trace WHERE thread_name != '' LIMIT 1; -1 --- enough rows (optimizations works "correctly") -SELECT count() > 100 FROM system.stack_trace; -1 diff --git a/tests/queries/0_stateless/01051_system_stack_trace.sql b/tests/queries/0_stateless/01051_system_stack_trace.sql deleted file mode 100644 index 93cc5763f19..00000000000 --- a/tests/queries/0_stateless/01051_system_stack_trace.sql +++ /dev/null @@ -1,24 +0,0 @@ --- Tags: no-parallel --- Tag no-parallel: to decrease failure probability of collecting stack traces - --- Process one thread at a time -SET max_block_size = 1; - --- It is OK to have bigger timeout here since: --- a) this test is marked as no-parallel --- b) there is a filter by thread_name, so it will send signals only to the threads with the name TCPHandler --- c) max_block_size is 1 -SET storage_system_stack_trace_pipe_read_timeout_ms = 5000; - --- { echo } -SELECT count() > 0 FROM system.stack_trace WHERE query_id != '' AND thread_name = 'TCPHandler'; --- opimization for not reading /proc/self/task/{}/comm and avoid sending signal -SELECT countIf(thread_id > 0) > 0 FROM system.stack_trace; --- optimization for trace -SELECT length(trace) > 0 FROM system.stack_trace WHERE length(trace) > 0 LIMIT 1; --- optimization for query_id -SELECT length(query_id) > 0 FROM system.stack_trace WHERE query_id != '' AND thread_name = 'TCPHandler' LIMIT 1; --- optimization for thread_name -SELECT length(thread_name) > 0 FROM system.stack_trace WHERE thread_name != '' LIMIT 1; --- enough rows (optimizations works "correctly") -SELECT count() > 100 FROM system.stack_trace; diff --git a/tests/queries/0_stateless/01070_alter_with_ttl.sql b/tests/queries/0_stateless/01070_alter_with_ttl.sql index 3adc3ccd6ae..44d422cbe6d 100644 --- a/tests/queries/0_stateless/01070_alter_with_ttl.sql +++ b/tests/queries/0_stateless/01070_alter_with_ttl.sql @@ -1,5 +1,7 @@ drop table if exists alter_ttl; +SET allow_suspicious_ttl_expressions = 1; + create table alter_ttl(i Int) engine = MergeTree order by i ttl toDate('2020-05-05'); alter table alter_ttl add column s String; alter table alter_ttl modify column s String ttl toDate('2020-01-01'); diff --git a/tests/queries/0_stateless/01070_materialize_ttl.sql b/tests/queries/0_stateless/01070_materialize_ttl.sql index b6a03f2ca10..b322b67882c 100644 --- a/tests/queries/0_stateless/01070_materialize_ttl.sql +++ b/tests/queries/0_stateless/01070_materialize_ttl.sql @@ -1,5 +1,7 @@ -- Tags: no-parallel +SET allow_suspicious_ttl_expressions = 1; + drop table if exists ttl; create table ttl (d Date, a Int) engine = MergeTree order by a partition by toDayOfMonth(d); diff --git a/tests/queries/0_stateless/01070_modify_ttl.sql b/tests/queries/0_stateless/01070_modify_ttl.sql index 0716ccd7043..4ffd59fc8a9 100644 --- a/tests/queries/0_stateless/01070_modify_ttl.sql +++ b/tests/queries/0_stateless/01070_modify_ttl.sql @@ -1,5 +1,7 @@ -- Tags: no-parallel +SET allow_suspicious_ttl_expressions = 1; + drop table if exists ttl; create table ttl (d Date, a Int) engine = MergeTree order by a partition by toDayOfMonth(d); diff --git a/tests/queries/0_stateless/01070_modify_ttl_recalc_only.sql b/tests/queries/0_stateless/01070_modify_ttl_recalc_only.sql index 7ac70d41871..2700cc03ff5 100644 --- a/tests/queries/0_stateless/01070_modify_ttl_recalc_only.sql +++ b/tests/queries/0_stateless/01070_modify_ttl_recalc_only.sql @@ -5,6 +5,8 @@ set mutations_sync = 2; -- system.parts has server default, timezone cannot be randomized set session_timezone = ''; +SET allow_suspicious_ttl_expressions = 1; + drop table if exists ttl; create table ttl (d Date, a Int) engine = MergeTree order by a partition by toDayOfMonth(d) diff --git a/tests/queries/0_stateless/01070_mutations_with_dependencies.sql b/tests/queries/0_stateless/01070_mutations_with_dependencies.sql index 566bb16b10c..813ebf3f5a7 100644 --- a/tests/queries/0_stateless/01070_mutations_with_dependencies.sql +++ b/tests/queries/0_stateless/01070_mutations_with_dependencies.sql @@ -34,6 +34,8 @@ select count() from ttl where s = 'b'; drop table ttl; -- check only that it doesn't throw exceptions. +SET allow_suspicious_ttl_expressions = 1; + create table ttl (i Int, s String) engine = MergeTree order by i ttl toDate('2000-01-01') TO DISK 'default'; alter table ttl materialize ttl; drop table ttl; diff --git a/tests/queries/0_stateless/01600_parts_states_metrics_long.reference b/tests/queries/0_stateless/01600_parts_states_metrics_long.reference deleted file mode 100644 index 98fb6a68656..00000000000 --- a/tests/queries/0_stateless/01600_parts_states_metrics_long.reference +++ /dev/null @@ -1,4 +0,0 @@ -1 -1 -1 -1 diff --git a/tests/queries/0_stateless/01600_parts_states_metrics_long.sh b/tests/queries/0_stateless/01600_parts_states_metrics_long.sh deleted file mode 100755 index 89ce84f6dbc..00000000000 --- a/tests/queries/0_stateless/01600_parts_states_metrics_long.sh +++ /dev/null @@ -1,40 +0,0 @@ -#!/usr/bin/env bash - -CURDIR=$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd) -# shellcheck source=../shell_config.sh -. "$CURDIR"/../shell_config.sh - -# NOTE: database = $CLICKHOUSE_DATABASE is unwanted -verify_sql="SELECT - (SELECT sumIf(value, metric = 'PartsActive'), sumIf(value, metric = 'PartsOutdated') FROM system.metrics) - = (SELECT sum(active), sum(NOT active) FROM - (SELECT active FROM system.parts UNION ALL SELECT active FROM system.projection_parts))" - -# The query is not atomic - it can compare states between system.parts and system.metrics from different points in time. -# So, there is inherent race condition. But it should get expected result eventually. -# In case of test failure, this code will do infinite loop and timeout. -verify() -{ - while true - do - result=$( $CLICKHOUSE_CLIENT -m --query="$verify_sql" ) - [ "$result" = "1" ] && break - sleep 0.1 - done - echo 1 -} - -$CLICKHOUSE_CLIENT --database_atomic_wait_for_drop_and_detach_synchronously=1 --query="DROP TABLE IF EXISTS test_table" -$CLICKHOUSE_CLIENT --query="CREATE TABLE test_table(data Date) ENGINE = MergeTree PARTITION BY toYear(data) ORDER BY data;" - -$CLICKHOUSE_CLIENT --query="INSERT INTO test_table VALUES ('1992-01-01')" -verify - -$CLICKHOUSE_CLIENT --query="INSERT INTO test_table VALUES ('1992-01-02')" -verify - -$CLICKHOUSE_CLIENT --query="OPTIMIZE TABLE test_table FINAL" -verify - -$CLICKHOUSE_CLIENT --database_atomic_wait_for_drop_and_detach_synchronously=1 --query="DROP TABLE test_table" -verify diff --git a/tests/queries/0_stateless/02000_join_on_const.reference b/tests/queries/0_stateless/02000_join_on_const.reference index e9d1c685fdd..848ecedf9e3 100644 --- a/tests/queries/0_stateless/02000_join_on_const.reference +++ b/tests/queries/0_stateless/02000_join_on_const.reference @@ -65,3 +65,7 @@ SELECT * FROM (SELECT 1 as a) as t1 FULL JOIN ( SELECT ('b', 256) as b ) AS t2 SELECT * FROM (SELECT 1 as a) as t1 SEMI JOIN ( SELECT ('b', 256) as b ) AS t2 ON NULL; SELECT * FROM (SELECT 1 as a) as t1 ANTI JOIN ( SELECT ('b', 256) as b ) AS t2 ON NULL; 1 ('',0) +2 +4 2 Nullable(UInt64) UInt8 +4 2 UInt64 Nullable(UInt8) +4 2 Nullable(UInt64) Nullable(UInt8) diff --git a/tests/queries/0_stateless/02000_join_on_const.sql b/tests/queries/0_stateless/02000_join_on_const.sql index 3205c084672..a68e75443d8 100644 --- a/tests/queries/0_stateless/02000_join_on_const.sql +++ b/tests/queries/0_stateless/02000_join_on_const.sql @@ -90,6 +90,31 @@ SELECT * FROM (SELECT 1 as a) as t1 ANTI JOIN ( SELECT ('b', 256) as b ) AS t2 -- { echoOff } +SELECT a + 1 +FROM (SELECT 1 as x) as t1 +LEFT JOIN ( SELECT 1 AS a ) AS t2 +ON TRUE +SETTINGS allow_experimental_analyzer=1, join_use_nulls=1; + +SELECT a + 1, x + 1, toTypeName(a), toTypeName(x) +FROM (SELECT 1 as x) as t1 +LEFT JOIN ( SELECT sum(number) as a from numbers(3) GROUP BY NULL) AS t2 +ON TRUE +SETTINGS allow_experimental_analyzer=1, join_use_nulls=1; + +SELECT a + 1, x + 1, toTypeName(a), toTypeName(x) +FROM (SELECT 1 as x) as t1 +RIGHT JOIN ( SELECT sum(number) as a from numbers(3) GROUP BY NULL) AS t2 +ON TRUE +SETTINGS allow_experimental_analyzer=1, join_use_nulls=1; + +SELECT a + 1, x + 1, toTypeName(a), toTypeName(x) +FROM (SELECT 1 as x) as t1 +FULL JOIN ( SELECT sum(number) as a from numbers(3) GROUP BY NULL) AS t2 +ON TRUE +SETTINGS allow_experimental_analyzer=1, join_use_nulls=1; + + DROP TABLE IF EXISTS t1; DROP TABLE IF EXISTS t2; diff --git a/tests/queries/0_stateless/02012_sha512_fixedstring.reference b/tests/queries/0_stateless/02012_sha512_fixedstring.reference index dfc2f87087c..93303a23fe1 100644 --- a/tests/queries/0_stateless/02012_sha512_fixedstring.reference +++ b/tests/queries/0_stateless/02012_sha512_fixedstring.reference @@ -1,5 +1,7 @@ CF83E1357EEFB8BDF1542850D66D8007D620E4050B5715DC83F4A921D36CE9CE47D0D13C5D85F2B0FF8318D2877EEC2F63B931BD47417A81A538327AF927DA3E DDAF35A193617ABACC417349AE20413112E6FA4E89A97EA20A9EEEE64B55D39A2192992A274FC1A836BA3C23A3FEEBBD454D4423643CE80E2A9AC94FA54CA49F +C672B8D1EF56ED28AB87C3622C5114069BDD3AD7B8F9737498D0C01ECEF0967A +53048E2681941EF99B2E29B76B4C7DABE4C2D0C634FC6D46E0E2F13107E7AF23 5809F3ECB4AA006F71AF562D4381F2BF64EA0931FD530E939740D0C38F6EEB2A71FA0113A21C170569D8319B8C4DE8A1C1A5ABA1A1C5B23A886B06712D373B9E 6FDB5E5BCCBD093ECC48DD262A99E6B867D6F48E1DAE014D26428365E7529B0022F000CBF852BEA38F43A2034E8FE7555AC41B9EA9E27FE72F4E968926998EA8 8018978D8AAE19322205E4CACFA045CDF7A0C4A5773A93FD24331064AFC5726F324B76802AA6FC30DFC412A6E5C3EEF4693AE4E2D0A1EA24A2D3EC46439B7923 @@ -20,3 +22,23 @@ E7A061D9B066E2CA44CF959A76FC04D8B02998CB9D46A60C19E015EA9389F3F9595CBBC4CC46E131 3438D75650E1EDB8A11EF4F63A5DFF239A70B28B6A14F13FCFDD14D02BE8BD00E84DF956C159CFDC85D6E44DB62D00E9206F40453FFD9CC97C38449527D33FF6 DB229C3A53B0340E94EFDA4D03B54F161313699757CAC312F377B731AE6C62010E0C0010E78F73E6D6B0BB438F644D176244B2614897799F9FA3F85DA980C218 FDD9FD54050D95855B8E3A34F3A54E309E1CA87CD44A8506EB10051D1CA650DB64ABD0BE4F4F44E45F630C22CA270FA7694AC2261DF2EFD766B8CED53F285A27 +99B13DC7BAF988EC34949E59E636884214C0155DA3BE69605F9DBEE0F9BB26FD +C93FBA93074FDEF8301B66D8DB835BFA2D50616EA108EC32CF617A7A40AF9746 +03CBA6DD6F5410B8E3E18D5271CAA786B2E67D8BE8281C893886B677C6A6F28D +BB612013EC3039A5943037FE9D92D976F85E8B69A3729019D674581DF101181A +40AC3D10E1820DBB473DC04CD32CCC14294F7BE5B3298C3BCC88F7F6AD7FCEB0 +659A9B47C2EC80125CB790FFCB1E234F2D33A93F6CF86DEF501305977A03BC75 +C03E2022A819F90B11FE1686C2827F67138B4DE7677C14D08972055336958263 +801A5F66EE0CD5EB1C214A9E61C3C4558A6795262E5A70D22AA392087F5C019D +AAD6E69EA25EECBC0A26CB60F9E2EBA878637A7CAA0CDF325D29C2516CE9D1A6 +B0B7F2E5E00E1F44578540A0B34138133654CA4803C245834D8C720BC46E6ADD +6475E517346B8C888418E1610E28969D10583CF431D05916A7DEE49633FE6279 +43A8BFDFBA86848376580714584F699E9775B7B31DDB7C4BBF0530D97564F5F9 +3B10D6BC94B31128583716432192F1B60D7BC23282A099F343CD5BB79323CD48 +B23EC2FC5B9D07E1853D50372600C8AEBA77F571B364D2D7D05E7B47ABA3E679 +F0DAA6078297F475B7E3B3968A2227A645DAA2ABA7426700532DCE1A04B67DB0 +4084B1624E1D9F5F80EE65C216316269E14A42CDF8607A651DFC602319CF9C66 +B63FBA87B2DCB490EE45B39F660A9DEFBD5383774B1C81D7A8CF9DE4C3FAA190 +BC70EB3AFD6D8252C92BB655920FFDCCC8239E99476697F1CDD4F847533671AC +6388E69B35B7A80469CD4C4EB49568C3143FEE6143112009120088A8607CC676 +7CE6CDE5DDE81925E714717F01771325298381B1934B021590965B30D23EC1C9 diff --git a/tests/queries/0_stateless/02012_sha512_fixedstring.sql b/tests/queries/0_stateless/02012_sha512_fixedstring.sql index cd014a58e65..ca9520350db 100644 --- a/tests/queries/0_stateless/02012_sha512_fixedstring.sql +++ b/tests/queries/0_stateless/02012_sha512_fixedstring.sql @@ -4,6 +4,9 @@ SELECT hex(SHA512('')); SELECT hex(SHA512('abc')); +SELECT hex(SHA512_256('')); +SELECT hex(SHA512_256('abc')); + DROP TABLE IF EXISTS defaults; CREATE TABLE defaults ( @@ -13,5 +16,6 @@ CREATE TABLE defaults INSERT INTO defaults SELECT s FROM generateRandom('s FixedString(20)', 1, 1, 1) LIMIT 20; SELECT hex(SHA512(s)) FROM defaults; +SELECT hex(SHA512_256(s)) FROM defaults; DROP TABLE defaults; diff --git a/tests/queries/0_stateless/02041_openssl_hash_functions_test.reference b/tests/queries/0_stateless/02041_openssl_hash_functions_test.reference index e5298179e6f..948e7913c5f 100644 --- a/tests/queries/0_stateless/02041_openssl_hash_functions_test.reference +++ b/tests/queries/0_stateless/02041_openssl_hash_functions_test.reference @@ -6,3 +6,4 @@ A94A8FE5CCB19BA61C4C0873D391E987982FBBD3 9F86D081884C7D659A2FEAA0C55AD015A3BF4F1B2B0B822CD15D6C15B0F00A08 768412320F7B0AA5812FCE428DC4706B3CAE50E02A64CAA16A782249BFE8EFC4B7EF1CCB126255D196047DFEDF17A0A9 EE26B0DD4AF7E749AA1A8EE3C10AE9923F618980772E473F8819A5D4940E0DB27AC185F8A0E1D5F84F88BC887FD67B143732C304CC5FA9AD8E6F57F50028A8FF +3D37FE58435E0D87323DEE4A2C1B339EF954DE63716EE79F5747F94D974F913F diff --git a/tests/queries/0_stateless/02041_openssl_hash_functions_test.sql b/tests/queries/0_stateless/02041_openssl_hash_functions_test.sql index 71940b24af4..7c712cf35fb 100644 --- a/tests/queries/0_stateless/02041_openssl_hash_functions_test.sql +++ b/tests/queries/0_stateless/02041_openssl_hash_functions_test.sql @@ -8,3 +8,4 @@ SELECT hex(SHA224('test')); SELECT hex(SHA256('test')); SELECT hex(SHA384('test')); SELECT hex(SHA512('test')); +SELECT hex(SHA512_256('test')); diff --git a/tests/queries/0_stateless/02184_hash_functions_and_ip_types.reference b/tests/queries/0_stateless/02184_hash_functions_and_ip_types.reference index b305806cd08..aad21567fca 100644 --- a/tests/queries/0_stateless/02184_hash_functions_and_ip_types.reference +++ b/tests/queries/0_stateless/02184_hash_functions_and_ip_types.reference @@ -34,6 +34,7 @@ hex(SHA1(ipv6)): A6D5DCE882AC44804382DE4639E6001612E1C8B5 hex(SHA224(ipv6)): F6995FD7BED2BCA21F68DAC6BBABE742DC1BA177BA8594CEF1715C52 hex(SHA256(ipv6)): F75497BAD6F7747BD6B150B6F69BA2DEE354F1C2A34B7BEA6183973B78640250 hex(SHA512(ipv6)): 0C2893CCBF44BC19CCF339AEED5B68CBFD5A2EF38263A48FE21C3379BA4438E7FF7A02F59D7542442C6E6ED538E6D13D65D3573DADB381651D3D8A5DEA232EAC +hex(SHA512_256(ipv6)): 1A2248FEB5A9D2D8D6C2482F132CFC19448B59DE75358E1F7ECAF444004F85A3 farmFingerprint64(ipv6): 6643158734288374888 javaHash(ipv6): 684606770 xxh3(ipv6): 4051340969481364358 diff --git a/tests/queries/0_stateless/02184_hash_functions_and_ip_types.sql b/tests/queries/0_stateless/02184_hash_functions_and_ip_types.sql index d96574ef4fe..c90c8b90c9e 100644 --- a/tests/queries/0_stateless/02184_hash_functions_and_ip_types.sql +++ b/tests/queries/0_stateless/02184_hash_functions_and_ip_types.sql @@ -38,6 +38,7 @@ SELECT hex(SHA224(ipv6)), hex(SHA256(ipv6)), hex(SHA512(ipv6)), + hex(SHA512_256(ipv6)), farmFingerprint64(ipv6), javaHash(ipv6), xxh3(ipv6), diff --git a/tests/queries/0_stateless/02296_ttl_non_deterministic.reference b/tests/queries/0_stateless/02296_ttl_non_deterministic.reference new file mode 100644 index 00000000000..e69de29bb2d diff --git a/tests/queries/0_stateless/02296_ttl_non_deterministic.sql b/tests/queries/0_stateless/02296_ttl_non_deterministic.sql new file mode 100644 index 00000000000..14d8979a624 --- /dev/null +++ b/tests/queries/0_stateless/02296_ttl_non_deterministic.sql @@ -0,0 +1,34 @@ +-- Tags: replica + +DROP TABLE IF EXISTS t_ttl_non_deterministic; + +CREATE TABLE t_ttl_non_deterministic(A Int64) +ENGINE = MergeTree ORDER BY A TTL now() + toIntervalMonth(1); -- {serverError BAD_ARGUMENTS} + +CREATE TABLE t_ttl_non_deterministic(A Int64) +ENGINE = ReplicatedMergeTree('/clickhouse/tables/{database}/ttl1', '1') ORDER BY A TTL now() + toIntervalMonth(1); -- {serverError BAD_ARGUMENTS} + + +CREATE TABLE t_ttl_non_deterministic(A Int64) ENGINE = MergeTree ORDER BY A; +ALTER TABLE t_ttl_non_deterministic MODIFY TTL now() + toIntervalMonth(1); -- {serverError BAD_ARGUMENTS} +DROP TABLE t_ttl_non_deterministic; + +CREATE TABLE t_ttl_non_deterministic(A Int64) ENGINE = ReplicatedMergeTree('/clickhouse/tables/{database}/ttl2', '1') ORDER BY A; +ALTER TABLE t_ttl_non_deterministic MODIFY TTL now() + toIntervalMonth(1); -- {serverError BAD_ARGUMENTS} +DROP TABLE t_ttl_non_deterministic; + + +CREATE TABLE t_ttl_non_deterministic(A Int64, B Int64 TTL now() + toIntervalMonth(1)) +ENGINE = MergeTree ORDER BY A; -- {serverError BAD_ARGUMENTS} + +CREATE TABLE t_ttl_non_deterministic(A Int64, B Int64 TTL now() + toIntervalMonth(1)) +ENGINE = ReplicatedMergeTree('/clickhouse/tables/{database}/ttl3', '1') ORDER BY A; -- {serverError BAD_ARGUMENTS} + + +CREATE TABLE t_ttl_non_deterministic(A Int64, B Int64) ENGINE = MergeTree ORDER BY A; +ALTER TABLE t_ttl_non_deterministic MODIFY COLUMN B Int64 TTL now() + toIntervalMonth(1); -- {serverError BAD_ARGUMENTS} +DROP TABLE t_ttl_non_deterministic; + +CREATE TABLE t_ttl_non_deterministic(A Int64, B Int64) ENGINE = ReplicatedMergeTree('/clickhouse/tables/{database}/ttl4', '1') ORDER BY A; +ALTER TABLE t_ttl_non_deterministic MODIFY COLUMN B Int64 TTL now() + toIntervalMonth(1); -- {serverError BAD_ARGUMENTS} +DROP TABLE t_ttl_non_deterministic; diff --git a/tests/queries/0_stateless/02789_jit_cannot_convert_column.reference b/tests/queries/0_stateless/02789_jit_cannot_convert_column.reference new file mode 100644 index 00000000000..e69de29bb2d diff --git a/tests/queries/0_stateless/02789_jit_cannot_convert_column.sql b/tests/queries/0_stateless/02789_jit_cannot_convert_column.sql new file mode 100644 index 00000000000..f5e694a38be --- /dev/null +++ b/tests/queries/0_stateless/02789_jit_cannot_convert_column.sql @@ -0,0 +1,11 @@ +SELECT + sum(c), + toInt32((h - null::Nullable(DateTime)) / 3600) + 1 AS a +FROM +( + SELECT count() AS c, h + FROM ( SELECT now() AS h ) + WHERE toInt32((h - null::Nullable(DateTime)) / 3600) + 1 = 1 + GROUP BY h +) +GROUP BY a settings min_count_to_compile_expression = 0; diff --git a/tests/queries/0_stateless/02864_statistic_exception.sql b/tests/queries/0_stateless/02864_statistic_exception.sql index c37f6b1ce06..092fa9bda85 100644 --- a/tests/queries/0_stateless/02864_statistic_exception.sql +++ b/tests/queries/0_stateless/02864_statistic_exception.sql @@ -47,7 +47,7 @@ ALTER TABLE t1 MATERIALIZE STATISTIC b TYPE tdigest; -- { serverError ILLEGAL_ST ALTER TABLE t1 ADD STATISTIC a TYPE tdigest; ALTER TABLE t1 ADD STATISTIC b TYPE tdigest; -ALTER TABLE t1 MODIFY COLUMN a Float64 TTL now() + INTERVAL 1 MONTH; +ALTER TABLE t1 MODIFY COLUMN a Float64 TTL toDateTime(b) + INTERVAL 1 MONTH; ALTER TABLE t1 MODIFY COLUMN a Int64; -- { serverError ALTER_OF_COLUMN_IS_FORBIDDEN } DROP TABLE t1; diff --git a/tests/queries/0_stateless/02896_memory_accounting_for_user.sh b/tests/queries/0_stateless/02896_memory_accounting_for_user.sh deleted file mode 100755 index f3016671420..00000000000 --- a/tests/queries/0_stateless/02896_memory_accounting_for_user.sh +++ /dev/null @@ -1,46 +0,0 @@ -#!/usr/bin/env bash -# Tags: no-parallel, long, no-random-settings - -CUR_DIR=$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd) -# shellcheck source=../shell_config.sh -. "$CUR_DIR"/../shell_config.sh - - -total_iterations=16 -parallelism=32 - -$CLICKHOUSE_CLIENT --query='DROP TABLE IF EXISTS test_inserts' -$CLICKHOUSE_CLIENT --query='CREATE TABLE test_inserts ENGINE=Null AS system.numbers' - -run_query() { - ( $CLICKHOUSE_CLIENT --query='SELECT * FROM numbers_mt(1000000) FORMAT CSV' | $CLICKHOUSE_CLIENT --max_threads 8 --max_memory_usage_for_user 1073741824 -q 'INSERT INTO test_inserts FORMAT CSV' 2>/dev/null ) -} - -for ((i = 1; i <= total_iterations; i++)); do - for ((j = 1; j <= parallelism; j++)); do - run_query & pids+=($!) - done - - EXIT_CODE=0 - new_pids=() - for pid in "${pids[@]:0:parallelism}"; do - CODE=0 - wait "${pid}" || CODE=$? - run_query & new_pids+=($!) - if [[ "${CODE}" != "0" ]]; then - EXIT_CODE=1; - fi - done - for pid in "${pids[@]:parallelism}"; do - CODE=0 - wait "${pid}" || CODE=$? - if [[ "${CODE}" != "0" ]]; then - EXIT_CODE=1; - fi - done - pids=("${new_pids[@]}") - - if [[ $EXIT_CODE -ne 0 ]]; then - exit $EXIT_CODE - fi -done diff --git a/tests/queries/0_stateless/02933_local_system_setting.reference b/tests/queries/0_stateless/02933_local_system_setting.reference new file mode 100644 index 00000000000..e69de29bb2d diff --git a/tests/queries/0_stateless/02933_local_system_setting.sh b/tests/queries/0_stateless/02933_local_system_setting.sh new file mode 100755 index 00000000000..c6d19f2445f --- /dev/null +++ b/tests/queries/0_stateless/02933_local_system_setting.sh @@ -0,0 +1,7 @@ +#!/usr/bin/env bash + +CURDIR=$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd) +# shellcheck source=../shell_config.sh +. "$CURDIR"/../shell_config.sh + +$CLICKHOUSE_LOCAL -q "select * from system.server_settings format Null;" diff --git a/tests/queries/0_stateless/02935_date_trunc_case_unsensitiveness.reference b/tests/queries/0_stateless/02935_date_trunc_case_unsensitiveness.reference new file mode 100644 index 00000000000..c8e3d37f08b --- /dev/null +++ b/tests/queries/0_stateless/02935_date_trunc_case_unsensitiveness.reference @@ -0,0 +1,9 @@ +2022-03-01 00:00:00 +2022-03-01 +2022-02-28 +2022-03-01 00:00:00 +2022-03-01 +2022-02-28 +2022-03-01 00:00:00 +2022-03-01 +2022-02-28 diff --git a/tests/queries/0_stateless/02935_date_trunc_case_unsensitiveness.sql b/tests/queries/0_stateless/02935_date_trunc_case_unsensitiveness.sql new file mode 100644 index 00000000000..ecf6877d477 --- /dev/null +++ b/tests/queries/0_stateless/02935_date_trunc_case_unsensitiveness.sql @@ -0,0 +1,9 @@ +SELECT dateTrunc('DAY', toDateTime('2022-03-01 12:55:55')); +SELECT dateTrunc('MONTH', toDateTime64('2022-03-01 12:55:55', 2)); +SELECT dateTrunc('WEEK', toDate('2022-03-01')); +SELECT dateTrunc('Day', toDateTime('2022-03-01 12:55:55')); +SELECT dateTrunc('Month', toDateTime64('2022-03-01 12:55:55', 2)); +SELECT dateTrunc('Week', toDate('2022-03-01')); +SELECT dateTrunc('day', toDateTime('2022-03-01 12:55:55')); +SELECT dateTrunc('month', toDateTime64('2022-03-01 12:55:55', 2)); +SELECT dateTrunc('week', toDate('2022-03-01')); diff --git a/tests/queries/0_stateless/02935_http_content_type_with_http_headers_progress.reference b/tests/queries/0_stateless/02935_http_content_type_with_http_headers_progress.reference new file mode 100644 index 00000000000..d847d4e22ef --- /dev/null +++ b/tests/queries/0_stateless/02935_http_content_type_with_http_headers_progress.reference @@ -0,0 +1,24 @@ +TSV + Content-Type: text/tab-separated-values; charset=UTF-8 +TabSeparatedWithNamesAndTypes + Content-Type: text/tab-separated-values; charset=UTF-8 +CSV + Content-Type: text/csv; charset=UTF-8; header=absent +CSVWithNames + Content-Type: text/csv; charset=UTF-8; header=present +Null + Content-Type: text/plain; charset=UTF-8 +Native + Content-Type: application/octet-stream +RowBinary + Content-Type: application/octet-stream +JSONStrings + Content-Type: application/json; charset=UTF-8 +JSON + Content-Type: application/json; charset=UTF-8 +JSONEachRow + Content-Type: application/x-ndjson; charset=UTF-8 +Values + Content-Type: text/plain; charset=UTF-8 +Vertical + Content-Type: text/plain; charset=UTF-8 diff --git a/tests/queries/0_stateless/02935_http_content_type_with_http_headers_progress.sh b/tests/queries/0_stateless/02935_http_content_type_with_http_headers_progress.sh new file mode 100755 index 00000000000..cd705650bad --- /dev/null +++ b/tests/queries/0_stateless/02935_http_content_type_with_http_headers_progress.sh @@ -0,0 +1,13 @@ +#!/usr/bin/env bash + +CURDIR=$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd) +# shellcheck source=../shell_config.sh +. "$CURDIR"/../shell_config.sh + +for frmt in TSV TabSeparatedWithNamesAndTypes CSV CSVWithNames Null Native RowBinary JSONStrings JSON JSONEachRow Values Vertical +do + echo $frmt + url="${CLICKHOUSE_URL}/?http_headers_progress_interval_ms=1&send_progress_in_http_headers=true&query=select+sleepEachRow(0.01)from+numbers(10)+FORMAT+${frmt}" + (seq 1 200| xargs -n1 -P0 -Ixxx curl -Ss -v -o /dev/null ${url} 2>&1|grep -Eo " Content-Type:.*$")|strings|sort -u +done + diff --git a/tests/queries/0_stateless/02941_any_RESPECT_NULL_sparse_column.reference b/tests/queries/0_stateless/02941_any_RESPECT_NULL_sparse_column.reference new file mode 100644 index 00000000000..573541ac970 --- /dev/null +++ b/tests/queries/0_stateless/02941_any_RESPECT_NULL_sparse_column.reference @@ -0,0 +1 @@ +0 diff --git a/tests/queries/0_stateless/02941_any_RESPECT_NULL_sparse_column.sql b/tests/queries/0_stateless/02941_any_RESPECT_NULL_sparse_column.sql new file mode 100644 index 00000000000..df86b740c21 --- /dev/null +++ b/tests/queries/0_stateless/02941_any_RESPECT_NULL_sparse_column.sql @@ -0,0 +1,5 @@ +-- regression for the case when aggregate function will be called with from==to for sparse column +DROP TABLE IF EXISTS data_sparse_column; +CREATE TABLE data_sparse_column (`key` Int64, `value` Int32) ENGINE = MergeTree ORDER BY key; +INSERT INTO data_sparse_column VALUES (1, 0); +SELECT any(value) RESPECT NULLS FROM data_sparse_column; diff --git a/tests/queries/1_stateful/00180_no_seek_avoiding_when_reading_from_cache.reference b/tests/queries/1_stateful/00180_no_seek_avoiding_when_reading_from_cache.reference deleted file mode 100644 index d05b1f927f4..00000000000 --- a/tests/queries/1_stateful/00180_no_seek_avoiding_when_reading_from_cache.reference +++ /dev/null @@ -1 +0,0 @@ -0 0 diff --git a/tests/queries/1_stateful/00180_no_seek_avoiding_when_reading_from_cache.sh b/tests/queries/1_stateful/00180_no_seek_avoiding_when_reading_from_cache.sh deleted file mode 100755 index 2e1b807c496..00000000000 --- a/tests/queries/1_stateful/00180_no_seek_avoiding_when_reading_from_cache.sh +++ /dev/null @@ -1,36 +0,0 @@ -#!/usr/bin/env bash - -# Tags: no-parallel, no-random-settings, long - -CURDIR=$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd) -# shellcheck source=../shell_config.sh -. "$CURDIR"/../shell_config.sh - - -# Test assumes that the whole table is residing in the cache, but `hits_s3` has only 128Mi of cache. -# So we need to create a smaller table. -$CLICKHOUSE_CLIENT -q "DROP TABLE IF EXISTS hits_s3_sampled" -$CLICKHOUSE_CLIENT -q "CREATE TABLE hits_s3_sampled AS test.hits_s3" -$CLICKHOUSE_CLIENT -q "INSERT INTO hits_s3_sampled SELECT * FROM test.hits_s3 SAMPLE 0.01" -$CLICKHOUSE_CLIENT -q "OPTIMIZE TABLE hits_s3_sampled FINAL" - -$CLICKHOUSE_CLIENT -q "SYSTEM DROP FILESYSTEM CACHE" - -# Warm up the cache -$CLICKHOUSE_CLIENT -q "SELECT * FROM hits_s3_sampled WHERE URL LIKE '%google%' ORDER BY EventTime LIMIT 10 FORMAT Null" -$CLICKHOUSE_CLIENT -q "SELECT * FROM hits_s3_sampled WHERE URL LIKE '%google%' ORDER BY EventTime LIMIT 10 FORMAT Null" - -query_id=02906_read_from_cache_$RANDOM -$CLICKHOUSE_CLIENT --query_id ${query_id} -q "SELECT * FROM hits_s3_sampled WHERE URL LIKE '%google%' ORDER BY EventTime LIMIT 10 FORMAT Null" - -$CLICKHOUSE_CLIENT -nq " - SYSTEM FLUSH LOGS; - - -- AsynchronousReaderIgnoredBytes = 0: no seek-avoiding happened - -- CachedReadBufferReadFromSourceBytes = 0: sanity check to ensure we read only from cache - SELECT ProfileEvents['AsynchronousReaderIgnoredBytes'], ProfileEvents['CachedReadBufferReadFromSourceBytes'] - FROM system.query_log - WHERE query_id = '$query_id' AND type = 'QueryFinish' AND event_date >= yesterday() AND current_database = currentDatabase() -" - -$CLICKHOUSE_CLIENT -q "DROP TABLE IF EXISTS hits_s3_sampled"