thevar1able/ClickHouse
1
0
Fork 0
mirror of https://github.com/ClickHouse/ClickHouse.git synced 2024-11-06 07:32:27 +00:00
Code Issues Packages Projects Releases Wiki Activity

Revert "Different approach to fix modulo"

Browse Source 
This reverts commit 665f1aa0f6.
This commit is contained in:
kssenii 2021-05-21 16:14:01 +00:00
parent 30845a383f
commit 94b5763dbd
16 changed files with 91 additions and 46 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

15
src/Storages/KeyDescription.cpp
View File

@ -2,7 +2,6 @@
#include <Functions/IFunction.h>
#include <Parsers/ASTIdentifier.h>
#include <Parsers/ASTFunction.h>
#include <Interpreters/ExpressionActions.h>
#include <Interpreters/ExpressionAnalyzer.h>
#include <Interpreters/TreeRewriter.h>
@ -87,25 +86,28 @@ KeyDescription KeyDescription::getKeyFromAST(
return getSortingKeyFromAST(definition_ast, columns, context, {});
}
void KeyDescription::moduloToModuloLegacyRecursive(ASTPtr node_expr)
bool KeyDescription::moduloToModuloLegacyRecursive(ASTPtr node_expr)
{
if (!node_expr)
return;
auto * function_expr = node_expr->as<ASTFunction>();
bool modulo_in_ast = false;
if (function_expr)
{
if (function_expr->name == "modulo")
{
function_expr->name = "moduloLegacy";
modulo_in_ast = true;
}
if (function_expr->arguments)
{
auto children = function_expr->arguments->children;
for (const auto & child : children)
moduloToModuloLegacyRecursive(child);
modulo_in_ast |= moduloToModuloLegacyRecursive(child);
}
}
return modulo_in_ast;
}
KeyDescription KeyDescription::getSortingKeyFromAST(
@ -115,11 +117,8 @@ KeyDescription KeyDescription::getSortingKeyFromAST(
const std::optional<String> & additional_column)
{
KeyDescription result;
ASTPtr adjusted_ast = definition_ast;
moduloToModuloLegacyRecursive(adjusted_ast);
result.definition_ast = adjusted_ast;
result.expression_list_ast = extractKeyExpressionList(adjusted_ast);
result.definition_ast = definition_ast;
result.expression_list_ast = extractKeyExpressionList(definition_ast);
if (additional_column)
{

8
src/Storages/MergeTree/DataPartsExchange.cpp
View File

@ -379,6 +379,7 @@ MergeTreeData::DataPartPtr Service::findPart(const String & name)
MergeTreeData::MutableDataPartPtr Fetcher::fetchPart(
const StorageMetadataPtr & metadata_snapshot,
ContextPtr context,
const String & part_name,
const String & replica_path,
const String & host,
@ -493,7 +494,7 @@ MergeTreeData::MutableDataPartPtr Fetcher::fetchPart(
if (e.code() != ErrorCodes::S3_ERROR)
throw;
/// Try again but without S3 copy
return fetchPart(metadata_snapshot, part_name, replica_path, host, port, timeouts,
return fetchPart(metadata_snapshot, context, part_name, replica_path, host, port, timeouts,
user, password, interserver_scheme, to_detached, tmp_prefix_, nullptr, false);
}
}
@ -557,7 +558,7 @@ MergeTreeData::MutableDataPartPtr Fetcher::fetchPart(
MergeTreeData::DataPart::Checksums checksums;
return part_type == "InMemory"
? downloadPartToMemory(part_name, part_uuid, metadata_snapshot, std::move(reservation), in, projections)
? downloadPartToMemory(part_name, part_uuid, metadata_snapshot, context, std::move(reservation), in, projections)
: downloadPartToDisk(part_name, replica_path, to_detached, tmp_prefix_, sync, reservation->getDisk(), in, projections, checksums);
}
@ -565,6 +566,7 @@ MergeTreeData::MutableDataPartPtr Fetcher::downloadPartToMemory(
const String & part_name,
const UUID & part_uuid,
const StorageMetadataPtr & metadata_snapshot,
ContextPtr context,
ReservationPtr reservation,
PooledReadWriteBufferFromHTTP & in,
size_t projections)
@ -619,7 +621,7 @@ MergeTreeData::MutableDataPartPtr Fetcher::downloadPartToMemory(
new_data_part->is_temp = true;
new_data_part->setColumns(block.getNamesAndTypesList());
new_data_part->minmax_idx.update(block, data.getMinMaxColumnsNames(metadata_snapshot->getPartitionKey()));
new_data_part->partition.create(metadata_snapshot, block, 0);
new_data_part->partition.create(metadata_snapshot, block, 0, context);
MergedBlockOutputStream part_out(
new_data_part, metadata_snapshot, block.getNamesAndTypesList(), {}, CompressionCodecFactory::instance().get("NONE", {}));

2
src/Storages/MergeTree/DataPartsExchange.h
View File

@ -65,6 +65,7 @@ public:
/// Downloads a part to tmp_directory. If to_detached - downloads to the `detached` directory.
MergeTreeData::MutableDataPartPtr fetchPart(
const StorageMetadataPtr & metadata_snapshot,
ContextPtr context,
const String & part_name,
const String & replica_path,
const String & host,
@ -106,6 +107,7 @@ private:
const String & part_name,
const UUID & part_uuid,
const StorageMetadataPtr & metadata_snapshot,
ContextPtr context,
ReservationPtr reservation,
PooledReadWriteBufferFromHTTP & in,
size_t projections);

2
src/Storages/MergeTree/MergeTreeBlockOutputStream.cpp
View File

@ -23,7 +23,7 @@ void MergeTreeBlockOutputStream::writePrefix()
void MergeTreeBlockOutputStream::write(const Block & block)
{
auto part_blocks = storage.writer.splitBlockIntoParts(block, max_parts_per_block, metadata_snapshot);
auto part_blocks = storage.writer.splitBlockIntoParts(block, max_parts_per_block, metadata_snapshot, context);
for (auto & current_block : part_blocks)
{
Stopwatch watch;

6
src/Storages/MergeTree/MergeTreeData.cpp
View File

@ -772,7 +772,7 @@ std::optional<UInt64> MergeTreeData::totalRowsByPartitionPredicateImpl(
// Generate valid expressions for filtering
bool valid = VirtualColumnUtils::prepareFilterBlockWithQuery(query_info.query, local_context, virtual_columns_block, expression_ast);
PartitionPruner partition_pruner(metadata_snapshot->getPartitionKey(), query_info, local_context, true /* strict */);
PartitionPruner partition_pruner(metadata_snapshot, query_info, local_context, true /* strict */);
if (partition_pruner.isUseless() && !valid)
return {};
@ -876,13 +876,13 @@ void MergeTreeData::loadDataParts(bool skip_sanity_checks)
{
/// Create and correctly initialize global WAL object
write_ahead_log = std::make_shared<MergeTreeWriteAheadLog>(*this, disk_ptr, it->name());
for (auto && part : write_ahead_log->restore(metadata_snapshot))
for (auto && part : write_ahead_log->restore(metadata_snapshot, getContext()))
parts_from_wal.push_back(std::move(part));
}
else if (settings->in_memory_parts_enable_wal)
{
MergeTreeWriteAheadLog wal(*this, disk_ptr, it->name());
for (auto && part : wal.restore(metadata_snapshot))
for (auto && part : wal.restore(metadata_snapshot, getContext()))
parts_from_wal.push_back(std::move(part));
}
}

2
src/Storages/MergeTree/MergeTreeDataSelectExecutor.cpp
View File

@ -503,7 +503,7 @@ QueryPlanPtr MergeTreeDataSelectExecutor::readFromParts(
minmax_idx_condition.emplace(
query_info, context, minmax_columns_names, data.getMinMaxExpr(partition_key, ExpressionActionsSettings::fromContext(context)));
partition_pruner.emplace(metadata_snapshot_base->getPartitionKey(), query_info, context, false /* strict */);
partition_pruner.emplace(metadata_snapshot_base, query_info, context, false /* strict */);
if (settings.force_index_by_date && (minmax_idx_condition->alwaysUnknownOrTrue() && partition_pruner->isUseless()))
{

11
src/Storages/MergeTree/MergeTreeDataWriter.cpp
View File

@ -140,7 +140,8 @@ void updateTTL(
}
BlocksWithPartition MergeTreeDataWriter::splitBlockIntoParts(const Block & block, size_t max_parts, const StorageMetadataPtr & metadata_snapshot)
BlocksWithPartition MergeTreeDataWriter::splitBlockIntoParts(
const Block & block, size_t max_parts, const StorageMetadataPtr & metadata_snapshot, ContextPtr context)
{
BlocksWithPartition result;
if (!block || !block.rows())
@ -155,12 +156,12 @@ BlocksWithPartition MergeTreeDataWriter::splitBlockIntoParts(const Block & block
}
Block block_copy = block;
const auto & partition_key = metadata_snapshot->getPartitionKey();
partition_key.expression->execute(block_copy);
/// After expression execution partition key columns will be added to block_copy with names regarding partition function.
auto partition_key_sample_block = MergeTreePartition::executePartitionByExpression(metadata_snapshot, block_copy, context);
ColumnRawPtrs partition_columns;
partition_columns.reserve(partition_key.sample_block.columns());
for (const ColumnWithTypeAndName & element : partition_key.sample_block)
partition_columns.reserve(partition_key_sample_block.columns());
for (const ColumnWithTypeAndName & element : partition_key_sample_block)
partition_columns.emplace_back(block_copy.getByName(element.name).column.get());
PODArray<size_t> partition_num_to_first_row;

2
src/Storages/MergeTree/MergeTreeDataWriter.h
View File

@ -40,7 +40,7 @@ public:
* (split rows by partition)
* Works deterministically: if same block was passed, function will return same result in same order.
*/
static BlocksWithPartition splitBlockIntoParts(const Block & block, size_t max_parts, const StorageMetadataPtr & metadata_snapshot);
static BlocksWithPartition splitBlockIntoParts(const Block & block, size_t max_parts, const StorageMetadataPtr & metadata_snapshot, ContextPtr context);
/** All rows must correspond to same partition.
* Returns part with unique name starting with 'tmp_', yet not added to MergeTreeData.

47
src/Storages/MergeTree/MergeTreePartition.cpp
View File

@ -169,22 +169,55 @@ void MergeTreePartition::store(const Block & partition_key_sample, const DiskPtr
out->finalize();
}
void MergeTreePartition::create(const StorageMetadataPtr & metadata_snapshot, Block block, size_t row)
void MergeTreePartition::create(const StorageMetadataPtr & metadata_snapshot, Block block, size_t row, ContextPtr context)
{
if (!metadata_snapshot->hasPartitionKey())
return;
const auto & partition_key = metadata_snapshot->getPartitionKey();
partition_key.expression->execute(block);
size_t partition_columns_num = partition_key.sample_block.columns();
auto partition_key_sample_block = executePartitionByExpression(metadata_snapshot, block, context);
size_t partition_columns_num = partition_key_sample_block.columns();
value.resize(partition_columns_num);
const String modulo_legacy_function_name = "moduloLegacy";
for (size_t i = 0; i < partition_columns_num; ++i)
{
const auto & column_name = partition_key.sample_block.getByPosition(i).name;
const auto & partition_column = block.getByName(column_name).column;
partition_column->get(row, value[i]);
const auto & column_name = partition_key_sample_block.getByPosition(i).name;
auto & partition_column = block.getByName(column_name);
/// Executing partition_by expression adds new columns to passed block according to partition functions.
/// The block is passed by reference and is used afterwards. `moduloLegacy` needs to be substituted back
/// with just `modulo`, because it was a temporary substitution.
if (column_name.starts_with(modulo_legacy_function_name))
partition_column.name = "modulo" + partition_column.name.substr(modulo_legacy_function_name.size());
partition_column.column->get(row, value[i]);
}
}
Block MergeTreePartition::executePartitionByExpression(const StorageMetadataPtr & metadata_snapshot, Block & block, ContextPtr context)
{
auto adjusted_partition_key = adjustPartitionKey(metadata_snapshot, context);
adjusted_partition_key.expression->execute(block);
return adjusted_partition_key.sample_block;
}
KeyDescription MergeTreePartition::adjustPartitionKey(const StorageMetadataPtr & metadata_snapshot, ContextPtr context)
{
const auto & partition_key = metadata_snapshot->getPartitionKey();
if (!partition_key.definition_ast)
return partition_key;
ASTPtr ast_copy = partition_key.definition_ast->clone();
/// Implementation of modulo function was changed from 8bit result type to 16bit. For backward compatibility partition by expression is always
/// calculated according to previous version - `moduloLegacy`.
if (KeyDescription::moduloToModuloLegacyRecursive(ast_copy))
{
auto adjusted_partition_key = KeyDescription::getKeyFromAST(ast_copy, metadata_snapshot->columns, context);
return adjusted_partition_key;
}
return partition_key;
}
}

8
src/Storages/MergeTree/MergeTreePartition.h
View File

@ -41,7 +41,13 @@ public:
void assign(const MergeTreePartition & other) { value.assign(other.value); }
void create(const StorageMetadataPtr & metadata_snapshot, Block block, size_t row);
void create(const StorageMetadataPtr & metadata_snapshot, Block block, size_t row, ContextPtr context);
/// Adjust partition key and execute its expression on block. Return sample block according to used expression.
static Block executePartitionByExpression(const StorageMetadataPtr & metadata_snapshot, Block & block, ContextPtr context);
/// Make a modified partition key with substitution from modulo to moduloLegacy. Used in paritionPruner.
static KeyDescription adjustPartitionKey(const StorageMetadataPtr & partition_key, ContextPtr context);
};
}

4
src/Storages/MergeTree/MergeTreeWriteAheadLog.cpp
View File

@ -111,7 +111,7 @@ void MergeTreeWriteAheadLog::rotate(const std::unique_lock<std::mutex> &)
init();
}
MergeTreeData::MutableDataPartsVector MergeTreeWriteAheadLog::restore(const StorageMetadataPtr & metadata_snapshot)
MergeTreeData::MutableDataPartsVector MergeTreeWriteAheadLog::restore(const StorageMetadataPtr & metadata_snapshot, ContextPtr context)
{
std::unique_lock lock(write_mutex);
@ -192,7 +192,7 @@ MergeTreeData::MutableDataPartsVector MergeTreeWriteAheadLog::restore(const Stor
MergedBlockOutputStream part_out(part, metadata_snapshot, block.getNamesAndTypesList(), {}, CompressionCodecFactory::instance().get("NONE", {}));
part->minmax_idx.update(block, storage.getMinMaxColumnsNames(metadata_snapshot->getPartitionKey()));
part->partition.create(metadata_snapshot, block, 0);
part->partition.create(metadata_snapshot, block, 0, context);
if (metadata_snapshot->hasSortingKey())
metadata_snapshot->getSortingKey().expression->execute(block);

2
src/Storages/MergeTree/MergeTreeWriteAheadLog.h
View File

@ -62,7 +62,7 @@ public:
void addPart(DataPartInMemoryPtr & part);
void dropPart(const String & part_name);
std::vector<MergeTreeMutableDataPartPtr> restore(const StorageMetadataPtr & metadata_snapshot);
std::vector<MergeTreeMutableDataPartPtr> restore(const StorageMetadataPtr & metadata_snapshot, ContextPtr context);
using MinMaxBlockNumber = std::pair<Int64, Int64>;
static std::optional<MinMaxBlockNumber> tryParseMinMaxBlockNumber(const String & filename);

9
src/Storages/MergeTree/PartitionPruner.h
View File

@ -14,15 +14,18 @@ class PartitionPruner
{
private:
std::unordered_map<String, bool> partition_filter_map;
const KeyDescription & partition_key;
/// partition_key is adjusted here (with substitution from modulo to moduloLegacy).
KeyDescription partition_key;
KeyCondition partition_condition;
bool useless;
using DataPart = IMergeTreeDataPart;
using DataPartPtr = std::shared_ptr<const DataPart>;
public:
PartitionPruner(const KeyDescription & partition_key_, const SelectQueryInfo & query_info, ContextPtr context, bool strict)
: partition_key(partition_key_)
PartitionPruner(const StorageMetadataPtr & metadata, const SelectQueryInfo & query_info, ContextPtr context, bool strict)
: partition_key(MergeTreePartition::adjustPartitionKey(metadata, context))
, partition_condition(
query_info, context, partition_key.column_names, partition_key.expression, true /* single_point */, strict)
, useless(strict ? partition_condition.anyUnknownOrAlwaysTrue() : partition_condition.alwaysUnknownOrTrue())

2
src/Storages/MergeTree/ReplicatedMergeTreeBlockOutputStream.cpp
View File

@ -136,7 +136,7 @@ void ReplicatedMergeTreeBlockOutputStream::write(const Block & block)
if (quorum)
checkQuorumPrecondition(zookeeper);
auto part_blocks = storage.writer.splitBlockIntoParts(block, max_parts_per_block, metadata_snapshot);
auto part_blocks = storage.writer.splitBlockIntoParts(block, max_parts_per_block, metadata_snapshot, context);
for (auto & current_block : part_blocks)
{

12
src/Storages/MergeTree/ReplicatedMergeTreeTableMetadata.cpp
View File

@ -206,18 +206,16 @@ void ReplicatedMergeTreeTableMetadata::checkImmutableFieldsEquals(const Replicat
", local: " + DB::toString(data_format_version.toUnderType()),
ErrorCodes::METADATA_MISMATCH);
//if (partition_key != from_zk.partition_key)
// throw Exception(
// "Existing table metadata in ZooKeeper differs in partition key expression."
// " Stored in ZooKeeper: " + from_zk.partition_key + ", local: " + partition_key,
// ErrorCodes::METADATA_MISMATCH);
if (partition_key != from_zk.partition_key)
throw Exception(
"Existing table metadata in ZooKeeper differs in partition key expression."
" Stored in ZooKeeper: " + from_zk.partition_key + ", local: " + partition_key,
ErrorCodes::METADATA_MISMATCH);
}
void ReplicatedMergeTreeTableMetadata::checkEquals(const ReplicatedMergeTreeTableMetadata & from_zk, const ColumnsDescription & columns, ContextPtr context) const
{
std::cerr << "\n\nKSSENII partitition key: " << partition_key << ", in zk: " << from_zk.partition_key << std::endl;
//String parsed = IndicesDescription::parse(from_zk.partition_key, columns, context).toString();
checkImmutableFieldsEquals(from_zk);

5
src/Storages/StorageReplicatedMergeTree.cpp
View File

@ -2466,7 +2466,7 @@ bool StorageReplicatedMergeTree::executeReplaceRange(const LogEntry & entry)
throw Exception("Interserver schemas are different '" + interserver_scheme + "' != '" + address.scheme + "', can't fetch part from " + address.host, ErrorCodes::LOGICAL_ERROR);
part_desc->res_part = fetcher.fetchPart(
metadata_snapshot, part_desc->found_new_part_name, source_replica_path,
metadata_snapshot, getContext(), part_desc->found_new_part_name, source_replica_path,
address.host, address.replication_port, timeouts, credentials->getUser(), credentials->getPassword(), interserver_scheme, false, TMP_PREFIX + "fetch_");
/// TODO: check columns_version of fetched part
@ -3929,6 +3929,7 @@ bool StorageReplicatedMergeTree::fetchPart(const String & part_name, const Stora
return fetcher.fetchPart(
metadata_snapshot,
getContext(),
part_name,
source_replica_path,
address.host,
@ -4084,7 +4085,7 @@ bool StorageReplicatedMergeTree::fetchExistsPart(const String & part_name, const
ErrorCodes::INTERSERVER_SCHEME_DOESNT_MATCH);
return fetcher.fetchPart(
metadata_snapshot, part_name, source_replica_path,
metadata_snapshot, getContext(), part_name, source_replica_path,
address.host, address.replication_port,
timeouts, credentials->getUser(), credentials->getPassword(), interserver_scheme, false, "", nullptr, true,
replaced_disk);