[Fix](delete) Fix delete job timeout when executing delete from ... #37363
Merged
dataroaring merged 3 commits intoapache:masterfrom Jul 6, 2024
Merged
[Fix](delete) Fix delete job timeout when executing delete from ... #37363dataroaring merged 3 commits intoapache:masterfrom
delete from ... #37363dataroaring merged 3 commits intoapache:masterfrom
Conversation
|
Thank you for your contribution to Apache Doris. Since 2024-03-18, the Document has been moved to doris-website. |
Contributor
Author
|
run buildall |
bobhan1
changed the title
delete from ... delete from ...
Contributor
|
clang-tidy review says "All clean, LGTM! 👍" |
1 similar comment
Contributor
|
clang-tidy review says "All clean, LGTM! 👍" |
Contributor
|
run buildall |
github-actions
bot
added
the
approved
Contributor
|
PR approved by at least one committer and no changes requested. |
Contributor
|
PR approved by anyone and no changes requested. |
bobhan1
changed the title
delete from ... delete from ...
TPC-H: Total hot run time: 40016 ms |
TPC-DS: Total hot run time: 174526 ms |
ClickBench: Total hot run time: 30.25 s |
Contributor
Author
|
run external |
Contributor
Author
|
run cloud_p1 |
bobhan1
force-pushed
the
fix-delete-timeout
branch
from
ae828d4 to
8fed24e
Compare
Contributor
Author
|
run buildall |
Contributor
|
clang-tidy review says "All clean, LGTM! 👍" |
TPC-H: Total hot run time: 39709 ms |
TPC-DS: Total hot run time: 171308 ms |
ClickBench: Total hot run time: 30.23 s |
Contributor
Author
|
run p0 |
Yukang-Lian
approved these changes
Jul 6, 2024
dataroaring
pushed a commit
that referenced
this pull request
Jul 17, 2024
…#37363) ## Proposed changes Currently, when FE execute delete job, it will send `REALTIME_PUSH` task to all affected replicas and **wait for all asynchronous tasks sent to the backend to return successful status results** or until timeout(which is at least 30s for delete job). If some replica failed to do the job and report an error for the task to FE, FE will retry the task to that replica. However, for some errors like `DELETE_INVALID_CONDITION`/`DELETE_INVALID_PARAMETERS`, we should fail and abort the delete job on FE directly and report the errors to users, rather than keep retrying in vain. So this PR let the delete job fail and abort directly on FE and report the errors to users if FE receives an above error from BE.
Closed
dataroaring
pushed a commit
that referenced
this pull request
Aug 28, 2024
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 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.
Proposed changes
Currently, when FE execute delete job, it will send
REALTIME_PUSHtask to all affected replicas and wait for all asynchronous tasks sent to the backend to return successful status results or until timeout(which is at least 30s for delete job). If some replica failed to do the job and report an error for the task to FE, FE will retry the task to that replica. However, for some errors likeDELETE_INVALID_CONDITION/DELETE_INVALID_PARAMETERS, we should fail and abort the delete job on FE directly and report the errors to users, rather than keep retrying in vain.So this PR let the delete job fail and abort directly on FE and report the errors to users if FE receives an above error from BE.
branch-2.1-pick: #37374