[
https://issues.apache.org/jira/browse/SOLR-16531?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17675736#comment-17675736
]
Jason Gerlowski edited comment on SOLR-16531 at 1/11/23 10:22 PM:
------------------------------------------------------------------
(Writing this right in a bit of a rush, hope to add/clarify later this evening.)
The Jan 11th deadline (as I understood it at least) was to come up with a
solution that could help us avoid a revert, and only revert if we didn't have
that in time. But at this point it looks like we do have that solution. So
I'd say: let's vet this thing. If it fixes the performance, great. If it
doesn't, well then we'll know the solution wasn't enough and we can revisit the
revert at that point based on what the remaining delta is. That seems most in
the spirit of the original agreement and should get everything resolved (and
everyone unblocked) quickly. I think we all want to move forward here!
In terms of vetting: I'll writeup some instructions or a script to run a little
solr-bench comparison, so that other people can provide data points and to
standardize our testing. (Maybe this'll spark some solr-bench interest and
improvements!) Of course, [PR|https://github.com/apache/solr/pull/1286] code
review also needed.
bq. I've prioritized my work schedule around this deadline such that I'm able
to start vetting other major commits
I hate that this has blocked other perf work that you've had in mind. If you
can expand a little bit on the nature of that "block", and why something like
the v2 feature flag doesn't help, maybe it's something we could work through
together?
was (Author: gerlowskija):
(Writing this right in a bit of a rush, hope to add/clarify later this evening.)
The Jan 11th deadline (as I understood it at least) was to come up with a
solution that could help us avoid a revert, and only revert if we didn't have
that in time. But at this point it looks like we do have that solution. So
I'd say: let's vet this thing. If it fixes the performance, great. If it
doesn't, well then we'll know the solution wasn't enough and we can revisit the
revert at that point based on what the remaining delta is. That seems most in
the spirit of the original agreement.
In terms of vetting: I'll writeup some instructions or a script to run a little
solr-bench comparison, so that other people can provide data points and to
standardize our testing. (Maybe this'll spark some solr-bench interest and
improvements!) Of course, [PR|https://github.com/apache/solr/pull/1286] code
review also needed.
bq. I've prioritized my work schedule around this deadline such that I'm able
to start vetting other major commits
I hate that this has blocked other perf work that you've had in mind. If you
can expand a little bit on the nature of that "block", and why something like
the v2 feature flag doesn't help, maybe it's something we could work through
together?
> Performance degradation due to introduction of JAX-RS
> -----------------------------------------------------
>
> Key: SOLR-16531
> URL: https://issues.apache.org/jira/browse/SOLR-16531
> Project: Solr
> Issue Type: Bug
> Security Level: Public(Default Security Level. Issues are Public)
> Reporter: Ishan Chattopadhyaya
> Priority: Blocker
> Fix For: 9.2
>
> Attachments: Screenshot from 2022-11-09 11-20-44.png,
> jaxrs-integration-with-app-per-configset-flamegraph.html,
> results-with-patch.tar.gz, vanilla-jaxrs-integration-flamegraph.html
>
>
> During performance benchmarking on branch_9x, I observed a slowdown in
> restart performance since commits in SOLR-16347. See attached screenshot.
> CC [~gerlowskija].
> http://mostly.cool/cluster-test-with-patch.html
> The benchmark is here:
> https://github.com/fullstorydev/solr-bench/blob/ishan/repeatable-jenkins/suites/cluster-test.json.
> This suite was run after retro-actively applying the parallelStream patch
> from SOLR-16414:
> https://github.com/apache/solr/commit/b33161d0cdd976fc0c3dc78c4afafceb4db671cf.diff
>
> Effort to automate these benchmarks is WIP and tracked here: SOLR-16525.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
