[enhancement](compaction) optimizing memory usage for compaction #36492
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. |
Contributor
Author
|
run buildall |
Contributor
|
clang-tidy review says "All clean, LGTM! 👍" |
Contributor
Author
|
run buildall |
Contributor
|
clang-tidy review says "All clean, LGTM! 👍" |
Contributor
|
clang-tidy review says "All clean, LGTM! 👍" |
Contributor
|
clang-tidy review says "All clean, LGTM! 👍" |
Contributor
Author
|
run buildall |
Contributor
|
clang-tidy review says "All clean, LGTM! 👍" |
TPC-H: Total hot run time: 40420 ms |
TPC-DS: Total hot run time: 173148 ms |
ClickBench: Total hot run time: 30.31 s |
|
TeamCity be ut coverage result: |
Contributor
Author
|
run buildall |
Contributor
|
clang-tidy review says "All clean, LGTM! 👍" |
TPC-H: Total hot run time: 41006 ms |
|
TeamCity be ut coverage result: |
dataroaring
reviewed
Jun 21, 2024
be/src/olap/base_compaction.cpp
Outdated
| break; | ||
| } | ||
| } | ||
|
|
Contributor
There was a problem hiding this comment.
Should be a defer, because there are other places where rowsets are removed.
dataroaring
reviewed
Jun 21, 2024
be/src/common/config.cpp
Outdated
|
|
||
| DEFINE_Bool(enable_base_compaction_idle_sched, "true"); | ||
| DEFINE_mInt64(base_compaction_min_rowset_num, "5"); | ||
| DEFINE_mInt64(base_compaction_max_rowset_num, "20"); |
Contributor
There was a problem hiding this comment.
We shoud limit comlumn num, it is waynum * column_num, because segment columns consumes memory too.
Contributor
There was a problem hiding this comment.
According to rowset num, we shoud limit score?
dataroaring
reviewed
Jun 21, 2024
| way_num = num_segments(); | ||
| CHECK(way_num > 0); | ||
| } | ||
| return way_num; |
Contributor
Author
There was a problem hiding this comment.
The score contains an empty rowset, while the merge way does not include an empty rowset
dataroaring
reviewed
Jun 21, 2024
be/src/olap/compaction.cpp
Outdated
| _merge_rowsets_latency_timer = ADD_TIMER(_profile, "merge_rowsets_latency"); | ||
| } | ||
|
|
||
| int64_t Compaction::merge_way() { |
dataroaring
reviewed
Jun 21, 2024
| << tablet->tablet_id() << " way cnt: " << way_cnt; | ||
| return 4096 - 32; | ||
| } | ||
| int64_t block_mem_limit = config::compaction_memory_bytes_limit / way_cnt; |
Contributor
There was a problem hiding this comment.
There are meta consuming memory too.
Contributor
Author
There was a problem hiding this comment.
This PR temporarily does not calculate the memory consumption of metadata
dataroaring
reviewed
Jun 21, 2024
| src_rowset_readers, dst_rowset_writer, max_rows_per_segment, stats_output, | ||
| key_group_cluster_key_idxes)); | ||
| key_group_cluster_key_idxes, batch_size, &(tablet->sample_infos[i]))); | ||
| if (is_key) { |
Contributor
There was a problem hiding this comment.
When memlimit happens, we should adjust batch_size.
ClickBench: Total hot run time: 30.47 s |
gavinchou
reviewed
Jun 25, 2024
be/src/olap/base_compaction.cpp
Outdated
| void BaseCompaction::_filter_input_rowset() { | ||
| auto rs_iter = _input_rowsets.begin(); | ||
| int score = 0; | ||
| while (rs_iter != _input_rowsets.end()) { |
Contributor
Author
|
run buildall |
![github-actions[bot]](https://avatars.githubusercontent.com/in/15368?s=60&v=4){kind=small}
| // specific language governing permissions and limitations | ||
| // under the License. | ||
|
|
||
| #include "olap/base_compaction.h" |
Contributor
There was a problem hiding this comment.
warning: 'olap/base_compaction.h' file not found [clang-diagnostic-error]
#include "olap/base_compaction.h"
^
TPC-H: Total hot run time: 40077 ms |
TPC-DS: Total hot run time: 172010 ms |
ClickBench: Total hot run time: 30.53 s |
Contributor
Author
|
run buildall |
TPC-H: Total hot run time: 40180 ms |
TPC-DS: Total hot run time: 173175 ms |
ClickBench: Total hot run time: 30.65 s |
Contributor
Author
|
run p0 |
Contributor
|
PR approved by at least one committer and no changes requested. |
github-actions
bot
added
the
approved
Contributor
|
PR approved by anyone and no changes requested. |
gavinchou
approved these changes
Jun 27, 2024
dataroaring
added a commit
to dataroaring/incubator-doris
that referenced
this pull request
Jun 29, 2024
…ion (apache#36492)" This reverts commit 9990181.
Contributor
|
revert and recommit in #37099 |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 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.
1 Limit the number of rowsets participating in base compaction.
2 Estimate the appropriate batch size by sampling and analyzing the memory consumption of blocks.