thevar1able/ClickHouse
1
0
Fork 0
mirror of https://github.com/ClickHouse/ClickHouse.git synced 2024-11-26 09:32:01 +00:00
Code Issues Packages Projects Releases Wiki Activity

Merge branch 'master' of github.com:yandex/ClickHouse

Browse Source
This commit is contained in:
Ivan Blinkov 2018-10-09 16:19:18 +03:00
commit 013602c269
8 changed files with 46 additions and 8 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
dbms/src/Interpreters/InterpreterSystemQuery.cpp
View File

@ -204,7 +204,7 @@ BlockIO InterpreterSystemQuery::execute()
throw Exception("There is no " + query.target_database + "." + query.target_table + " replicated table",
ErrorCodes::BAD_ARGUMENTS);
break;
case Type::FLUSH_SYSTEM_TABLES:
case Type::FLUSH_LOGS:
executeCommandsAndThrowIfError(
[&] () { if (auto query_log = context.getQueryLog(false)) query_log->flush(); },
[&] () { if (auto part_log = context.getPartLog("", false)) part_log->flush(); },

4
dbms/src/Parsers/ASTSystemQuery.cpp
View File

@ -65,8 +65,8 @@ const char * ASTSystemQuery::typeToString(Type type)
return "STOP REPLICATION QUEUES";
case Type::START_REPLICATION_QUEUES:
return "START REPLICATION QUEUES";
case Type::FLUSH_SYSTEM_TABLES:
return "FLUSH SYSTEM TABLES";
case Type::FLUSH_LOGS:
return "FLUSH LOGS";
default:
throw Exception("Unknown SYSTEM query command", ErrorCodes::BAD_TYPE_OF_FIELD);
}

2
dbms/src/Parsers/ASTSystemQuery.h
View File

@ -39,7 +39,7 @@ public:
START_REPLICATEDS_SENDS,
STOP_REPLICATION_QUEUES,
START_REPLICATION_QUEUES,
FLUSH_SYSTEM_TABLES,
FLUSH_LOGS,
END
};

4
dbms/tests/queries/0_stateless/00634_performance_introspection_and_logging.sh
View File

@ -34,7 +34,7 @@ DROP TABLE test.null;"
heavy_cpu_query="SELECT ignore(sum(sipHash64(hex(sipHash64(hex(sipHash64(hex(number)))))))) FROM (SELECT * FROM system.numbers_mt LIMIT 1000000)"
$CLICKHOUSE_CLIENT $settings --max_threads=1 -q "$heavy_cpu_query"
$CLICKHOUSE_CLIENT $settings -q "SYSTEM FLUSH SYSTEM TABLES"
$CLICKHOUSE_CLIENT $settings -q "SYSTEM FLUSH LOGS"
$CLICKHOUSE_CLIENT $settings -q "
WITH
any(query_duration_ms*1000) AS duration,
@ -53,7 +53,7 @@ SELECT
# Check ProfileEvents in query_thread_log
$CLICKHOUSE_CLIENT $settings --max_threads=3 -q "$heavy_cpu_query"
$CLICKHOUSE_CLIENT $settings -q "SYSTEM FLUSH SYSTEM TABLES"
$CLICKHOUSE_CLIENT $settings -q "SYSTEM FLUSH LOGS"
query_id=`$CLICKHOUSE_CLIENT $settings -q "SELECT query_id FROM system.query_log WHERE event_date >= today()-1 AND type=2 AND query='$heavy_cpu_query' ORDER BY event_time DESC LIMIT 1"`
query_elapsed=`$CLICKHOUSE_CLIENT $settings -q "SELECT query_duration_ms*1000 FROM system.query_log WHERE event_date >= today()-1 AND type=2 AND query_id='$query_id' ORDER BY event_time DESC LIMIT 1"`
threads=`$CLICKHOUSE_CLIENT $settings -q "SELECT length(thread_numbers) FROM system.query_log WHERE event_date >= today()-1 AND type=2 AND query_id='$query_id' ORDER BY event_time DESC LIMIT 1"`

1
dbms/tests/queries/0_stateless/00731_merge_tree_select_opened_files.reference Normal file
View File

@ -0,0 +1 @@
4

28
dbms/tests/queries/0_stateless/00731_merge_tree_select_opened_files.sh Executable file
View File

@ -0,0 +1,28 @@
#!/usr/bin/env bash
set -e
CURDIR=$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)
. $CURDIR/../shell_config.sh
cur_name=${BASH_SOURCE[0]}
settings="$server_logs --log_queries=1 --log_query_threads=1 --log_profile_events=1 --log_query_settings=1"
# Test insert logging on each block and checkPacket() method
$CLICKHOUSE_CLIENT $settings -n -q "
DROP TABLE IF EXISTS test.merge_tree_table;
CREATE TABLE test.merge_tree_table (id UInt64, date Date, uid UInt32) ENGINE = MergeTree(date, id, 8192);"
$CLICKHOUSE_CLIENT $settings -q "INSERT INTO test.merge_tree_table SELECT (intHash64(number)) % 10000, toDate('2018-08-01'), rand() FROM system.numbers LIMIT 10000000;"
$CLICKHOUSE_CLIENT $settings -q "OPTIMIZE TABLE test.merge_tree_table FINAL;"
toching_many_parts_query="SELECT count() from (SELECT toDayOfWeek(date) as m, id, count() FROM test.merge_tree_table GROUP BY id, m ORDER BY count() DESC LIMIT 10 SETTINGS max_threads = 1)"
$CLICKHOUSE_CLIENT $settings -q "$toching_many_parts_query" &> /dev/null
$CLICKHOUSE_CLIENT $settings -q "SYSTEM FLUSH LOGS"
$CLICKHOUSE_CLIENT $settings -q "SELECT pi.Values FROM system.query_log ARRAY JOIN ProfileEvents as pi WHERE query='$toching_many_parts_query' and pi.Names = 'FileOpen' ORDER BY event_time DESC LIMIT 1;"
$CLICKHOUSE_CLIENT $settings -q "DROP TABLE IF EXISTS test.merge_tree_table;"

2
docs/ru/operations/table_engines/dictionary.md
View File

@ -2,7 +2,7 @@
# Dictionary
Движок `Dictionary` отображает данные словаря как таблицу ClickHouse.
Движок `Dictionary` отображает данные [словаря](../../query_language/dicts/external_dicts.md) как таблицу ClickHouse.
Рассмотрим для примера словарь `products` со следующей конфигурацией:

1
docs/zh/faq/general.md
View File

@ -1 +0,0 @@
../../en/faq/general.md

10
docs/zh/faq/general.md Normal file
View File

@ -0,0 +1,10 @@
# 常见问题
## 为什么不使用MapReduce之类的产品呢?
我们可以将MapReduce这类的系统称为分布式计算系统，其reduce操作基于分布式排序。其中最常见的开源解决方案是[Apache Hadoop](http://hadoop.apache.org)。 Yandex使用他们的内部解决方案YT。
这些系统不适合在线查询，因为它们的延迟高。换句话说，它们不能用作Web接口的后端服务。这些系统对于实时数据更新是没有用的。如果操作的结果和所有中间结果（如果有的话）位于单个服务器的内存中，则分布式排序不是执行reduce操作的最佳方式，但这通常是在线查询的情况。在这种情况下，哈希表是执行reduce操作的最佳方式。优化map-reduce任务的常用方法是使用内存中的哈希表进行预聚合（部分reduce），用户手动执行此优化操作。分布式排序是运行简单map-reduce任务时性能降低的主要原因之一。
大多数MapReduce系统允许您在集群上执行任意代码。但是，声明性查询语言更适合OLAP，以便快速运行实验。例如，Hadoop包含Hive和Pig，Cloudera Impala或Shark（过时）for Spark，以及Spark SQL、Presto和Apache Drill。与专业系统相比，运行此类任务时的性能非常不理想，所以将这些系统用作Web接口的后端服务是不现实的，因为延迟相对较高。