ClickHouse/src/Storages/MergeTree/MergeTreeSelectProcessor.cpp

128 lines
4.3 KiB
C++
Raw Normal View History

2019-10-01 16:50:08 +00:00
#include <Storages/MergeTree/MergeTreeSelectProcessor.h>
#include <Storages/MergeTree/MergeTreeBaseSelectProcessor.h>
2019-10-10 16:30:30 +00:00
#include <Storages/MergeTree/IMergeTreeReader.h>
#include <Interpreters/Context.h>
2016-11-20 12:43:20 +00:00
namespace DB
{
namespace ErrorCodes
{
extern const int MEMORY_LIMIT_EXCEEDED;
2016-11-20 12:43:20 +00:00
}
2019-10-01 16:50:08 +00:00
MergeTreeSelectProcessor::MergeTreeSelectProcessor(
2018-10-17 03:13:00 +00:00
const MergeTreeData & storage_,
const StorageMetadataPtr & metadata_snapshot_,
const MergeTreeData::DataPartPtr & owned_data_part_,
2019-02-10 16:55:12 +00:00
UInt64 max_block_size_rows_,
size_t preferred_block_size_bytes_,
size_t preferred_max_column_in_block_size_bytes_,
2019-07-19 14:56:00 +00:00
Names required_columns_,
2019-10-04 15:40:05 +00:00
MarkRanges mark_ranges_,
bool use_uncompressed_cache_,
const PrewhereInfoPtr & prewhere_info_,
2019-08-03 11:02:40 +00:00
bool check_columns_,
const MergeTreeReaderSettings & reader_settings_,
const Names & virt_column_names_,
2017-04-05 20:34:19 +00:00
size_t part_index_in_query_,
bool quiet)
:
2019-10-01 16:50:08 +00:00
MergeTreeBaseSelectProcessor{
metadata_snapshot_->getSampleBlockForColumns(required_columns_, storage_.getVirtuals()),
storage_, metadata_snapshot_, prewhere_info_, max_block_size_rows_,
2019-12-19 13:10:57 +00:00
preferred_block_size_bytes_, preferred_max_column_in_block_size_bytes_,
reader_settings_, use_uncompressed_cache_, virt_column_names_},
2019-10-01 16:50:08 +00:00
required_columns{std::move(required_columns_)},
2017-04-05 20:34:19 +00:00
data_part{owned_data_part_},
2019-10-04 15:40:05 +00:00
all_mark_ranges(std::move(mark_ranges_)),
2017-04-05 20:34:19 +00:00
part_index_in_query(part_index_in_query_),
2019-08-03 11:02:40 +00:00
check_columns(check_columns_),
path(data_part->getFullRelativePath())
{
/// Let's estimate total number of rows for progress bar.
for (const auto & range : all_mark_ranges)
2018-05-23 19:34:37 +00:00
total_marks_count += range.end - range.begin;
2019-07-15 20:26:55 +00:00
size_t total_rows = data_part->index_granularity.getRowsCountInRanges(all_mark_ranges);
if (!quiet)
2020-05-23 22:24:01 +00:00
LOG_TRACE(log, "Reading {} ranges from part {}, approx. {} rows starting from {}",
2020-05-23 21:41:35 +00:00
all_mark_ranges.size(), data_part->name, total_rows,
data_part->index_granularity.getMarkStartingRow(all_mark_ranges.front().begin));
2018-02-23 10:02:29 +00:00
addTotalRowsApprox(total_rows);
ordered_names = header_without_virtual_columns.getNames();
2018-04-16 12:21:36 +00:00
}
2019-10-01 16:50:08 +00:00
bool MergeTreeSelectProcessor::getNewTask()
try
2016-11-20 12:43:20 +00:00
{
2018-05-23 19:34:37 +00:00
/// Produce no more than one task
if (!is_first_task || total_marks_count == 0)
2017-04-05 20:34:19 +00:00
{
finish();
2017-04-05 20:34:19 +00:00
return false;
}
is_first_task = false;
2019-07-19 14:56:00 +00:00
task_columns = getReadTaskColumns(storage, data_part, required_columns, prewhere_info, check_columns);
2017-04-05 20:34:19 +00:00
2018-09-03 05:06:19 +00:00
auto size_predictor = (preferred_block_size_bytes == 0)
? nullptr
: std::make_unique<MergeTreeBlockSizePredictor>(data_part, ordered_names, data_part->storage.getSampleBlock());
2017-04-05 20:34:19 +00:00
2019-07-19 14:56:00 +00:00
/// will be used to distinguish between PREWHERE and WHERE columns when applying filter
const auto & column_names = task_columns.columns.getNames();
column_name_set = NameSet{column_names.begin(), column_names.end()};
task = std::make_unique<MergeTreeReadTask>(
data_part, all_mark_ranges, part_index_in_query, ordered_names, column_name_set, task_columns.columns,
2019-07-19 14:56:00 +00:00
task_columns.pre_columns, prewhere_info && prewhere_info->remove_prewhere_column,
task_columns.should_reorder, std::move(size_predictor));
if (!reader)
{
if (use_uncompressed_cache)
owned_uncompressed_cache = storage.global_context.getUncompressedCache();
owned_mark_cache = storage.global_context.getMarkCache();
2019-10-10 16:30:30 +00:00
reader = data_part->getReader(task_columns.columns, all_mark_ranges,
owned_uncompressed_cache.get(), owned_mark_cache.get(), reader_settings);
if (prewhere_info)
2019-10-10 16:30:30 +00:00
pre_reader = data_part->getReader(task_columns.pre_columns, all_mark_ranges,
owned_uncompressed_cache.get(), owned_mark_cache.get(), reader_settings);
}
2017-04-05 20:34:19 +00:00
return true;
}
catch (...)
2017-04-05 20:34:19 +00:00
{
/// Suspicion of the broken part. A part is added to the queue for verification.
if (getCurrentExceptionCode() != ErrorCodes::MEMORY_LIMIT_EXCEEDED)
storage.reportBrokenPart(data_part->name);
throw;
}
2019-10-01 16:50:08 +00:00
void MergeTreeSelectProcessor::finish()
{
/** Close the files (before destroying the object).
* When many sources are created, but simultaneously reading only a few of them,
* buffers don't waste memory.
*/
reader.reset();
pre_reader.reset();
data_part.reset();
2016-11-20 12:43:20 +00:00
}
2019-10-01 16:50:08 +00:00
MergeTreeSelectProcessor::~MergeTreeSelectProcessor() = default;
2016-11-20 12:43:20 +00:00
}