[RFC] XrdAdaptor: Use inactive sources first before trying to open a new source when handling a failed request #47593
Conversation
This file contains 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
|
cms-bot internal usage |
This was referenced Mar 13, 2025
|
+code-checks Logs: https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-47593/44088 |
ArturAkh
reviewed
Mar 13, 2025
| // similar logic as in checkSourcesImpl() | ||
| // assume the "sort open delay" doesn't matter in case of a request failure | ||
| auto bestInactiveSource = | ||
| std::min_element(m_inactiveSources.begin(), |
There was a problem hiding this comment.
Is the best quality really the lowest number?
There was a problem hiding this comment.
As far as I can tell, yes. See e.g. existing code
cmssw/Utilities/XrdAdaptor/src/XrdRequestManager.cc
Lines 455 to 460 in eb45085
There is some discussion about the quality metric in https://github.com/cms-sw/cmssw/blob/master/Utilities/XrdAdaptor/doc/multisource_algorithm_design.txt
and maybe it is computed in https://github.com/cms-sw/cmssw/blob/master/Utilities/XrdAdaptor/src/QualityMetric.cc, but I haven't tried to understand what is done there.
|
FYI @osschar |
|
@cmsbuild, please test |
|
+1 Size: This PR adds an extra 32KB to repository Comparison SummarySummary:
|
…r::requestFailure() According to Brian Bockelman this was the intention, but apparently was never coded.
makortel
force-pushed
the
xrdAdaptorRetryInactive
branch
from
27252f2 to
688b9dc
Compare
cmsbuild
added
tests-pending
code-checks-pending
and removed
tests-approved
code-checks-approved
labels
|
+code-checks Logs: https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-47593/44101 |
|
Pull request #47593 was updated. |
|
With @bockjoo's test #43162 (comment) the present state of this PR results in the following actions with sources (with increased verbosity of XrdAdaptor printouts with process.MessageLogger.cerr.threshold = "INFO"
process.MessageLogger.cerr.XrdAdaptorInternal = dict()
process.MessageLogger.cerr.XrdAdaptorLvl1 = dict()
process.MessageLogger.cerr.XrdAdaptorLvl3 = dict()) |
|
To summarize #47593 (comment)
At least this private test demonstrates the code in this PR has an impact. |
|
@cmsbuild, please test |
|
+1 Size: This PR adds an extra 32KB to repository Comparison SummarySummary:
|
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
PR description:
In the context of #43162 and #46086 (comment), this PR changes the XrdAdaptor behavior for request failures. Previously, after disabling the source that resulted the request failure, it passed on the request to another active source, and if there were no active sources, it tried to open a new source. This PR changes to logic to
The first commit generalizes the code where the failed source is removed from the active sources container. I found it confusing that the code checked only the first 2 active sources (even if there would only ever be at most 2 active sources). Admittedly, if there would be more than 2 active sources, the commit would change the behavior, but I would argue for the better.
Resolves cms-sw/framework-team#1306
PR validation:
Code compiles.
If this PR is a backport please specify the original PR and why you need to backport that PR. If this PR will be backported please specify to which release cycle the backport is meant for:
Possibly to be backported to 15_0_X.