Hi,

Tim, I fixed the test bug. It was caused by missing shards parameter. Actually 
no bug in distributed sorting, just a test bug introduced in the original PR.

Sorry for the failures: Failures happened when your server was too fast, not 
too slow 😊

Uwe

-----
Uwe Schindler
Achterdiek 19, D-28357 Bremen
https://www.thetaphi.de
eMail: [email protected]

> -----Original Message-----
> From: Timothy Potter <[email protected]>
> Sent: Friday, August 13, 2021 7:44 PM
> To: [email protected]
> Subject: Re: Consistent failure of DistributedQueryComponentCustomSortTest
> on branch_8x when running in CI env
> 
> I meant it is consistently failing in our CI ;-) Or at least I've not
> seen it pass since the update in our env ...
> 
> On Fri, Aug 13, 2021 at 3:38 AM Uwe Schindler <[email protected]> wrote:
> >
> > Hi,
> >
> > No, it is NOT consistent. It happens only seldom, depending on load of build
> server.
> >
> > But this is actually a good sign! The result merging code is broken, but 
> > this
> was just uncovered by the new test added in this commit. The problem is that
> the "partialResults" flag is not set on result if the last shard timed out 
> while
> sorting. I will look into this.
> >
> > I hit the problem locally one time, but I thougt it was caused by the test 
> > code
> written with SolrJ that checked the partialResults flag by string comparison
> instead of using the named list correctly. But I was wrong.
> >
> > So thanks to the NPE fix we found a bug in the merging code!
> >
> > Uwe
> >
> > -----
> > Uwe Schindler
> > Achterdiek 19, D-28357 Bremen
> > https://www.thetaphi.de
> > eMail: [email protected]
> >
> > > -----Original Message-----
> > > From: Timothy Potter <[email protected]>
> > > Sent: Thursday, August 12, 2021 10:31 PM
> > > To: [email protected]
> > > Subject: Re: Consistent failure of
> DistributedQueryComponentCustomSortTest
> > > on branch_8x when running in CI env
> > >
> > > Not just our CI ... same failure here:
> > > https://jenkins.thetaphi.de/job/Lucene-Solr-8.x-MacOSX/2184/
> > >
> > > On Thu, Aug 12, 2021 at 2:03 PM Timothy Potter <[email protected]>
> > > wrote:
> > > >
> > > > I'm seeing the following failure on branch_8x which looks like new
> > > > code added to the test in commit:
> > > > https://github.com/apache/lucene-
> > > solr/commit/d99980516bd3470e9e120428914d1270d5ddf91b
> > > >
> > > > It passes locally in my IDE but fails in an internal Jenkins server
> > > > running JDK 11 & 15:
> > > >
> > > > java.lang.AssertionError: Number of documents (1) is different from
> > > > number of expected values (3
> > > > at
> > >
> __randomizedtesting.SeedInfo.seed([FF67B6C9248EFA4D:773389138A7297B5]
> > > :0)
> > > > at org.junit.Assert.fail(Assert.java:89)
> > > > at
> org.apache.solr.SolrTestCaseJ4.assertFieldValues(SolrTestCaseJ4.java:2184)
> > > > at
> > >
> org.apache.solr.handler.component.DistributedQueryComponentCustomSortTe
> > > st.test(DistributedQueryComponentCustomSortTest.java:128)
> > >
> > > ---------------------------------------------------------------------
> > > To unsubscribe, e-mail: [email protected]
> > > For additional commands, e-mail: [email protected]
> >
> >
> > ---------------------------------------------------------------------
> > To unsubscribe, e-mail: [email protected]
> > For additional commands, e-mail: [email protected]
> >
> 
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [email protected]
> For additional commands, e-mail: [email protected]


---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to