Uh oh!
There was an error while loading. Please reload this page.
[opt](split) get file splits in batch mode - #34032
Merged
Merged
Conversation
doris-robot
commented
Apr 23, 2024
Thank you for your contribution to Apache Doris. Since 2024-03-18, the Document has been moved to doris-website. |
Contributor
clang-tidy review says "All clean, LGTM! 👍" |
Uh oh!
There was an error while loading. Please reload this page.
Uh oh!
There was an error while loading. Please reload this page.
Uh oh!
There was an error while loading. Please reload this page.
Contributor
clang-tidy review says "All clean, LGTM! 👍" |
AshinGauforce-pushed
the
split_source
branch
2 times, most recently
from
May 7, 2024 07:34
9e436b6 to
380cea0CompareAshinGau
marked this pull request as ready for review
May 7, 2024 07:34
Contributor
clang-tidy review says "All clean, LGTM! 👍" |
1 similar comment
Contributor
clang-tidy review says "All clean, LGTM! 👍" |
AshinGau
commented
May 7, 2024
MemberAuthor
run buildall |
Contributor
clang-tidy review says "All clean, LGTM! 👍" |
AshinGau
commented
May 7, 2024
MemberAuthor
run buildall |
Contributor
clang-tidy review says "All clean, LGTM! 👍" |
AshinGau
commented
May 8, 2024
MemberAuthor
run buildall |
Contributor
clang-tidy review says "All clean, LGTM! 👍" |
doris-robot
commented
May 8, 2024
TeamCity be ut coverage result: |
Contributor
PR approved by at least one committer and no changes requested. |
Contributor
PR approved by anyone and no changes requested. |
| SplitSource splitSource = new SplitSource( | ||
| this::splitToScanRange, backend, locationProperties, splits, pathPartitionKeys); | ||
| splitSources.add(splitSource); | ||
| SplitSourceManager.registerSplitSource(splitSource); |
| import java.util.Map; | ||
| import java.util.concurrent.ConcurrentHashMap; | ||
| public class SplitSourceManager { |
Contributor
There was a problem hiding this comment.
Suggest:
- not using singletion
- extends
MasterDaemonclass
| return QeProcessorImpl.INSTANCE.reportExecStatus(params, getClientAddr()); | ||
| } | ||
| public TFetchSplitBatchResult fetchSplitBatch(TFetchSplitBatchRequest request) throws TException { |
| LOG(WARNING) << "Failed to get batch of split source: {}, try to reopen" << e1.what(); | ||
| RETURN_IF_ERROR(coord.reopen()); | ||
| try { | ||
| coord->fetchSplitBatch(result, request); |
Contributor
There was a problem hiding this comment.
Maybe we should not retry when failure.
If first call fail, it is highly possible the second would fail too.
Simply fail this query to avoid avalanche
| @ConfField(mutable = true, masterOnly = false, description = { | ||
| "如果切片数量超过阈值，BE将通过batch方式获取scan ranges", | ||
| "If the number of splits exceeds the threshold, scan ranges will be got through batch mode."}) | ||
| public static int num_splits_in_batch_mode = 10000; |
Contributor
There was a problem hiding this comment.
Better be a session varible?
doris-robot
commented
May 13, 2024
TPC-H: Total hot run time: 41933 ms |
doris-robot
commented
May 13, 2024
TPC-DS: Total hot run time: 187732 ms |
doris-robot
commented
May 14, 2024
TeamCity be ut coverage result: |
Contributor
PR approved by at least one committer and no changes requested. |
AshinGau added a commit
to AshinGau/incubator-doris
that referenced
this pull request
May 21, 2024
When scanning a table with many files, It will take a lot of time to transfer splits to backends.(20s of the following 1209172 splits). Therefore, using batch mode to fetch the file splits, BE can do scanning while fetch the file splits.
AshinGau added a commit
to AshinGau/incubator-doris
that referenced
this pull request
May 21, 2024
When scanning a table with many files, It will take a lot of time to transfer splits to backends.(20s of the following 1209172 splits). Therefore, using batch mode to fetch the file splits, BE can do scanning while fetch the file splits.
AshinGau added a commit
to AshinGau/incubator-doris
that referenced
this pull request
May 21, 2024
When scanning a table with many files, It will take a lot of time to transfer splits to backends.(20s of the following 1209172 splits). Therefore, using batch mode to fetch the file splits, BE can do scanning while fetch the file splits.
AshinGau added a commit
to AshinGau/incubator-doris
that referenced
this pull request
May 21, 2024
When scanning a table with many files, It will take a lot of time to transfer splits to backends.(20s of the following 1209172 splits). Therefore, using batch mode to fetch the file splits, BE can do scanning while fetch the file splits.
morningman pushed a commit
that referenced
this pull request
May 21, 2024
morningman added a commit
that referenced
this pull request
Aug 6, 2024
PR #34032 introduce a new method to get splits batch by batch, but it removed a logic that BE will merge scan ranges to avoid too many scan ranges being scheduled. This PR mainly changes: 1. Add scan range merging logic back. 2. Change the default file split size from 8MB to 64MB, to avoid too many small split.
morningman added a commit
to morningman/doris
that referenced
this pull request
Aug 6, 2024
PR apache#34032 introduce a new method to get splits batch by batch, but it removed a logic that BE will merge scan ranges to avoid too many scan ranges being scheduled. This PR mainly changes: 1. Add scan range merging logic back. 2. Change the default file split size from 8MB to 64MB, to avoid too many small split.
dataroaring pushed a commit
that referenced
this pull request
Aug 11, 2024
PR #34032 introduce a new method to get splits batch by batch, but it removed a logic that BE will merge scan ranges to avoid too many scan ranges being scheduled. This PR mainly changes: 1. Add scan range merging logic back. 2. Change the default file split size from 8MB to 64MB, to avoid too many small split.
dataroaring pushed a commit
that referenced
this pull request
Aug 16, 2024
PR #34032 introduce a new method to get splits batch by batch, but it removed a logic that BE will merge scan ranges to avoid too many scan ranges being scheduled. This PR mainly changes: 1. Add scan range merging logic back. 2. Change the default file split size from 8MB to 64MB, to avoid too many small split.
HappenLee pushed a commit
to HappenLee/incubator-doris
that referenced
this pull request
Apr 24, 2026
When scanning a table with many files, It will take a lot of time to transfer splits to backends.(20s of the following 1209172 splits). Therefore, using batch mode to fetch the file splits, BE can do scanning while fetch the file splits.
HappenLee pushed a commit
to HappenLee/incubator-doris
that referenced
this pull request
Apr 24, 2026
PR apache#34032 introduce a new method to get splits batch by batch, but it removed a logic that BE will merge scan ranges to avoid too many scan ranges being scheduled. This PR mainly changes: 1. Add scan range merging logic back. 2. Change the default file split size from 8MB to 64MB, to avoid too many small split.
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
Sign up for freeto 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
When scanning a table with many files, It will take a lot of time to transfer splits to backends.(20s of the following 1209172 splits).
Therefore, using batch mode to fetch the file splits, BE can do scanning while fetch the file splits.
Further comments
If this is a relatively large or complex change, kick off the discussion at dev@doris.apache.org by explaining why you chose the solution you did and what alternatives you considered, etc...