Thanks for the heads up, James! I haven't been able to reproduce TestLBHttp2SolrClient, unfortunately. Looking at the historical failure rates it has been flaky for some time - which assuages my concerns about a "new bug" a good bit. IMO it needn't be resolved prior to 9.7.1. (Though I intend to keep looking at it.)
On TestPullReplica the news is a bit better! Many of the failures there are related to an errant "AlreadyClosedException" that bubbles out of the IndexFetcher code. By happy coincidence Sanjay Dutt merged a fix for this last week, which I've been able to backport to branch_9_7. So in both cases I think we're unblocked again for a 9.7.1! Best, Jason On Mon, Nov 4, 2024 at 10:30 AM James Dyer <jdyer....@gmail.com> wrote: > > Jason, you may want to know that PR #2828 (to be merged to main soon) > will rename TestLBHttp2SolrClient to LBHttp2SolrClientIntegrationTest. > I did not look into why it fails sometimes, but I have a feeling it is > because it relies on Thread.sleep and clock time. > > On Fri, Nov 1, 2024 at 1:46 PM Jason Gerlowski <gerlowsk...@gmail.com> wrote: > > > > Related to the release - I'm noticing a lot more test failures than > > usual on fucit. A lot of this is noise on 'main' due to the recent > > JDK21 upgrade, and Curator introduction. But even filtering out > > 'main', there is a higher than usual failure rate for 9.x and > > branch_9_7. > > > > I'm going to start digging into a few of these, starting with > > TestLBHttp2SolrClient and TestPullReplica. I'd encourage other folks > > who have been making changes recently to grab a test failure or two as > > well. I know our builds have never been perfect, but hopefully we can > > improve the situation somewhat before anyone starts cutting RCs. > > > > Best, > > > > Jason > > > > On Fri, Nov 1, 2024 at 6:33 AM Eric Pugh <de...@yahoo.com.invalid> wrote: > > > > > > Good perspective Anshum. I may have been infected with 10x fever! > > > > > > > > > Sent from my iPhone > > > > > > > On Oct 30, 2024, at 12:43 PM, Anshum Gupta <ans...@anshumgupta.net> > > > > wrote: > > > > I agree, let's do both 9.7.1 and 9.8 as that is what makes sense for > > > > users. > > > > > > > > We also assumed 8.11.1 would be the last 8x release but it didn't really > > > > end up that way. Let's support 9x until it makes sense for us and the > > > > user > > > > community to do so. I personally see more 9x releases in the future. > > > > > > > > On Wed, Oct 30, 2024 at 4:48 AM Jan Høydahl <jan....@cominvent.com> > > > > wrote: > > > > > > > >> +1 to an 9.7.1 stability / bugfix release. > > > >> > > > >> And we should continue minor 9.x releases every couple months until we > > > >> are > > > >> ready to do a quality 10.0 release with JDK, Jetty, SolrJ etc sorted > > > >> out. > > > >> We of course hope that is soonish, but let's not decide which is the > > > >> lastest 9.x version until we are almost ready to present a 10.0 RC. > > > >> > > > >> Jan > > > >> > > > >>> 30. okt. 2024 kl. 11:25 skrev Jason Gerlowski <gerlowsk...@gmail.com>: > > > >>> Why not both? > > > >>> I think a 9.8 would be awesome to get some additional deprecations and > > > >>> minor features out. But we've stacked up a number of pretty serious > > > >>> 9.7 bugs that we should try to get fixed for users independent of > > > >>> that: SOLR-17515, SOLR-17464, SOLR-17457 (which prevents our starting > > > >>> on Windows altogether!), etc. So IMO we need a 9.7.1 as well. > > > >>> Better to have two separate releases each with their own timelines > > > >>> than to delay getting those 9.7 fixes out, or to rush 9.8 features and > > > >>> deprecations. I'm sure folks will have a lot they'll want to fit into > > > >>> a 9.8 if it's framed as the last minor release before 10. (which > > > >>> sounds like a reasonable assumption Eric's making above). > > > >>> Best, > > > >>> Jason > > > >>> On Tue, Oct 29, 2024 at 12:35 PM Gus Heck <gus.h...@gmail.com> wrote: > > > >>>> I'm targeting 9.8 for SOLR-17151 if the rest of my life would ever > > > >>>> stop > > > >>>> needing immediate attention. > > > >>>> On Tue, Oct 29, 2024 at 12:16 PM David Eric Pugh > > > >> <de...@yahoo.com.invalid> > > > >>>> wrote: > > > >>>>> Would it make more sense to do a 9.8? There are lots of CLI > > > >> deprecation > > > >>>>> in 9.8 that would be nice to see the light of day before 10.0 is > > > >> shipped! > > > >>>>> On Tuesday, October 29, 2024 at 09:58:18 AM EDT, Jan Høydahl < > > > >>>>> jan....@cominvent.com> wrote: > > > >>>>> Thanks Anshum. We have 11 bugs fixed for 9.8 < > > > >> https://issues.apache.org/jira/issues/?jql=project%20=%20SOLR%20AND%20issuetype%20=%20Bug%20AND%20status%20=%20Resolved%20AND%20fixVersion%20=%209.8 > > > >>>>> so far, may consider backporting some of them in 9.7.1. > > > >>>>> Jan > > > >>>>>> 28. okt. 2024 kl. 18:57 skrev Anshum Gupta <ans...@apache.org>: > > > >>>>>> Hi everyone, > > > >>>>>> With the basic-auth + recovery related issue, we'll need to release > > > >> 9.7.1 > > > >>>>>> soon. I'll be happy to be th release manager for the same whenever > > > >>>>>> the > > > >>>>>> changes are merged and we're ready for a release. > > > >>>>>> -Anshum > > > >>>> -- > > > >>>> http://www.needhamsoftware.com (work) > > > >>>> https://a.co/d/b2sZLD9 (my fantasy fiction book) > > > >>> --------------------------------------------------------------------- > > > >>> To unsubscribe, e-mail: dev-unsubscr...@solr.apache.org > > > >>> For additional commands, e-mail: dev-h...@solr.apache.org > > > >> > > > >> > > > >> --------------------------------------------------------------------- > > > >> To unsubscribe, e-mail: dev-unsubscr...@solr.apache.org > > > >> For additional commands, e-mail: dev-h...@solr.apache.org > > > > > > > > -- > > > > Anshum Gupta > > > > > > > > > --------------------------------------------------------------------- > > > To unsubscribe, e-mail: dev-unsubscr...@solr.apache.org > > > For additional commands, e-mail: dev-h...@solr.apache.org > > > > > > > --------------------------------------------------------------------- > > To unsubscribe, e-mail: dev-unsubscr...@solr.apache.org > > For additional commands, e-mail: dev-h...@solr.apache.org > > > > --------------------------------------------------------------------- > To unsubscribe, e-mail: dev-unsubscr...@solr.apache.org > For additional commands, e-mail: dev-h...@solr.apache.org > --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@solr.apache.org For additional commands, e-mail: dev-h...@solr.apache.org
