[Fix](multi-catalog) Fix core in orc and parquet reader sometimes after low mem exception. #36574
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
Thank you for your contribution to Apache Doris. Since 2024-03-18, the Document has been moved to doris-website. |
kaka11chen
force-pushed
the
fix_low_mem_core_in_orc_parquet_reader
branch
from
0b15d2a to
234af45
Compare
Contributor
|
clang-tidy review says "All clean, LGTM! 👍" |
1 similar comment
Contributor
|
clang-tidy review says "All clean, LGTM! 👍" |
Contributor
Author
|
run buildall |
morningman
force-pushed
the
fix_low_mem_core_in_orc_parquet_reader
branch
from
234af45 to
5926dd3
Compare
Contributor
|
run buildall |
Contributor
|
clang-tidy review says "All clean, LGTM! 👍" |
TPC-H: Total hot run time: 40291 ms |
TPC-DS: Total hot run time: 174994 ms |
ClickBench: Total hot run time: 30.71 s |
kaka11chen
force-pushed
the
fix_low_mem_core_in_orc_parquet_reader
branch
from
5926dd3 to
8083e36
Compare
Contributor
Author
|
run buildall |
Contributor
|
clang-tidy review says "All clean, LGTM! 👍" |
TPC-H: Total hot run time: 39538 ms |
TPC-DS: Total hot run time: 173519 ms |
ClickBench: Total hot run time: 31.51 s |
Contributor
Author
|
run p0 |
1 similar comment
Contributor
Author
|
run p0 |
Contributor
|
PR approved by at least one committer and no changes requested. |
github-actions
bot
added
approved
Contributor
|
PR approved by anyone and no changes requested. |
AshinGau
previously approved these changes
Jun 27, 2024
morningman
requested changes
Jul 1, 2024
Contributor
morningman
left a comment
morningman
left a comment
There was a problem hiding this comment.
2.1 has reverted this PR #37086
github-actions
bot
removed
the
approved
kaka11chen
force-pushed
the
fix_low_mem_core_in_orc_parquet_reader
branch
from
8083e36 to
1b27117
Compare
…er low mem exception.
kaka11chen
force-pushed
the
fix_low_mem_core_in_orc_parquet_reader
branch
from
1b27117 to
b6b50e5
Compare
Contributor
Author
|
run buildall |
Contributor
|
clang-tidy review says "All clean, LGTM! 👍" |
TPC-H: Total hot run time: 39759 ms |
TPC-DS: Total hot run time: 174441 ms |
ClickBench: Total hot run time: 30.33 s |
github-actions
bot
added
the
approved
Contributor
|
PR approved by at least one committer and no changes requested. |
zy-kkk
approved these changes
Jul 13, 2024
seawinde
pushed a commit
to seawinde/doris
that referenced
this pull request
Jul 17, 2024
…er low mem exception. (apache#36574) ## Proposed changes ### Issue ``` SIGSEGV address not mapped to object (@0x8) received by PID 2264997 (TID 2267559 OR 0x7ff60b2a4640) from PID 8; stack trace: *** 0# doris::signal::(anonymous namespace)::FailureSignalHandler(int, siginfo_t*, void*) at /home/zcp/repo_center/doris_branch-2.1/doris/be/src/common/signal_handler.h:421 1# PosixSignals::chained_handler(int, siginfo*, void*) [clone .part.0] in /usr/lib/jvm/java-17-openjdk-amd64/lib/server/libjvm.so 2# JVM_handle_linux_signal in /usr/lib/jvm/java-17-openjdk-amd64/lib/server/libjvm.so 3# 0x00007FFB96D77520 in /lib/x86_64-linux-gnu/libc.so.6 4# doris::vectorized::Block::clear_column_data(int) at /home/zcp/repo_center/doris_branch-2.1/doris/be/src/vec/core/block.cpp:705 5# doris::vectorized::OrcReader::get_next_block_impl(doris::vectorized::Block*, unsigned long*, bool*) in /mnt/hdd01/PERFORMANCE_ENV/be/lib/doris_be 6# doris::vectorized::OrcReader::get_next_block(doris::vectorized::Block*, unsigned long*, bool*) at /home/zcp/repo_center/doris_branch-2.1/doris/be/src/vec/exec/format/orc/vorc_reader.cpp:1533 7# doris::vectorized::VFileScanner::_get_block_wrapped(doris::RuntimeState*, doris::vectorized::Block*, bool*) at /home/zcp/repo_center/doris_branch-2.1/doris/be/src/vec/exec/scan/vfile_scanner.cpp:355 8# doris::vectorized::VFileScanner::_get_block_impl(doris::RuntimeState*, doris::vectorized::Block*, bool*) at /home/zcp/repo_center/doris_branch-2.1/doris/be/src/vec/exec/scan/vfile_scanner.cpp:298 9# doris::vectorized::VScanner::get_block(doris::RuntimeState*, doris::vectorized::Block*, bool*) in /mnt/hdd01/PERFORMANCE_ENV/be/lib/doris_be 10# doris::vectorized::VScanner::get_block_after_projects(doris::RuntimeState*, doris::vectorized::Block*, bool*) at /home/zcp/repo_center/doris_branch-2.1/doris/be/src/vec/exec/scan/vscanner.cpp:96 11# doris::vectorized::ScannerScheduler::_scanner_scan(std::shared_ptr<doris::vectorized::ScannerContext>, std::shared_ptr<doris::vectorized::ScanTask>) at /home/zcp/repo_center/doris_branch-2.1/doris/be/src/vec/exec/scan/scanner_scheduler.cpp:258 ``` ### Root cause It is found that when executing conjuncts expressions when there is insufficient memory, column names and types may be inserted, but the columns are not inserted into the block, which eventually leads to a crash. If an exception occurs after `block->insert({nullptr, _data_type, _expr_name});`, it will lead to the situation mentioned above ``` Status VectorizedFnCall::_do_execute(doris::vectorized::VExprContext* context, doris::vectorized::Block* block, int* result_column_id, std::vector<size_t>& args) { ... // prepare a column to save result block->insert({nullptr, _data_type, _expr_name}); if (_can_fast_execute) { auto can_fast_execute = fast_execute(*block, args, num_columns_without_result, block->rows(), _function->get_name()); if (can_fast_execute) { *result_column_id = num_columns_without_result; return Status::OK(); } } ``` ### Solution apache#37086
dataroaring
pushed a commit
that referenced
this pull request
Jul 17, 2024
…er low mem exception. (#36574) ## Proposed changes ### Issue ``` SIGSEGV address not mapped to object (@0x8) received by PID 2264997 (TID 2267559 OR 0x7ff60b2a4640) from PID 8; stack trace: *** 0# doris::signal::(anonymous namespace)::FailureSignalHandler(int, siginfo_t*, void*) at /home/zcp/repo_center/doris_branch-2.1/doris/be/src/common/signal_handler.h:421 1# PosixSignals::chained_handler(int, siginfo*, void*) [clone .part.0] in /usr/lib/jvm/java-17-openjdk-amd64/lib/server/libjvm.so 2# JVM_handle_linux_signal in /usr/lib/jvm/java-17-openjdk-amd64/lib/server/libjvm.so 3# 0x00007FFB96D77520 in /lib/x86_64-linux-gnu/libc.so.6 4# doris::vectorized::Block::clear_column_data(int) at /home/zcp/repo_center/doris_branch-2.1/doris/be/src/vec/core/block.cpp:705 5# doris::vectorized::OrcReader::get_next_block_impl(doris::vectorized::Block*, unsigned long*, bool*) in /mnt/hdd01/PERFORMANCE_ENV/be/lib/doris_be 6# doris::vectorized::OrcReader::get_next_block(doris::vectorized::Block*, unsigned long*, bool*) at /home/zcp/repo_center/doris_branch-2.1/doris/be/src/vec/exec/format/orc/vorc_reader.cpp:1533 7# doris::vectorized::VFileScanner::_get_block_wrapped(doris::RuntimeState*, doris::vectorized::Block*, bool*) at /home/zcp/repo_center/doris_branch-2.1/doris/be/src/vec/exec/scan/vfile_scanner.cpp:355 8# doris::vectorized::VFileScanner::_get_block_impl(doris::RuntimeState*, doris::vectorized::Block*, bool*) at /home/zcp/repo_center/doris_branch-2.1/doris/be/src/vec/exec/scan/vfile_scanner.cpp:298 9# doris::vectorized::VScanner::get_block(doris::RuntimeState*, doris::vectorized::Block*, bool*) in /mnt/hdd01/PERFORMANCE_ENV/be/lib/doris_be 10# doris::vectorized::VScanner::get_block_after_projects(doris::RuntimeState*, doris::vectorized::Block*, bool*) at /home/zcp/repo_center/doris_branch-2.1/doris/be/src/vec/exec/scan/vscanner.cpp:96 11# doris::vectorized::ScannerScheduler::_scanner_scan(std::shared_ptr<doris::vectorized::ScannerContext>, std::shared_ptr<doris::vectorized::ScanTask>) at /home/zcp/repo_center/doris_branch-2.1/doris/be/src/vec/exec/scan/scanner_scheduler.cpp:258 ``` ### Root cause It is found that when executing conjuncts expressions when there is insufficient memory, column names and types may be inserted, but the columns are not inserted into the block, which eventually leads to a crash. If an exception occurs after `block->insert({nullptr, _data_type, _expr_name});`, it will lead to the situation mentioned above ``` Status VectorizedFnCall::_do_execute(doris::vectorized::VExprContext* context, doris::vectorized::Block* block, int* result_column_id, std::vector<size_t>& args) { ... // prepare a column to save result block->insert({nullptr, _data_type, _expr_name}); if (_can_fast_execute) { auto can_fast_execute = fast_execute(*block, args, num_columns_without_result, block->rows(), _function->get_name()); if (can_fast_execute) { *result_column_id = num_columns_without_result; return Status::OK(); } } ``` ### Solution #37086
Closed
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
7 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Proposed changes
Issue
Root cause
It is found that when executing conjuncts expressions when there is insufficient memory, column names and types may be inserted, but the columns are not inserted into the block, which eventually leads to a crash.
If an exception occurs after
block->insert({nullptr, _data_type, _expr_name});, it will lead to the situation mentioned aboveSolution
#37086