mikemccand commented on pull request #1623:
URL: https://github.com/apache/lucene-solr/pull/1623#issuecomment-679088291
> @mikemccand this test reproduces on master too I opened
https://issues.apache.org/jira/browse/LUCENE-9477
Great, thanks @s1monw! Given that beasting is now
mikemccand commented on pull request #1623:
URL: https://github.com/apache/lucene-solr/pull/1623#issuecomment-678796112
Hmm, not that I am using the larger (`2000mb`) line docs file, and that
might be needed to provoke failure.
mikemccand commented on pull request #1623:
URL: https://github.com/apache/lucene-solr/pull/1623#issuecomment-678796020
> @mikemccand I think we are ready here WDYT?
+1! Thanks @s1monw!
I will continue beasting. The failures are very rare now ... I ran 196
iters of slow +
mikemccand commented on pull request #1623:
URL: https://github.com/apache/lucene-solr/pull/1623#issuecomment-677710206
I will restart beasting on the latest PR now. Thanks @s1monw!
This is an automated message from the
mikemccand commented on pull request #1623:
URL: https://github.com/apache/lucene-solr/pull/1623#issuecomment-67545
Those were the only two failures found after 1033 test iterations!
This is an automated message from the
mikemccand commented on pull request #1623:
URL: https://github.com/apache/lucene-solr/pull/1623#issuecomment-675459732
Another failure that reproduces only on the PR:
```
[junit4:pickseed] Seed property 'tests.seed' already defined:
CADC6D7945159855
[junit4] says jolly
mikemccand commented on pull request #1623:
URL: https://github.com/apache/lucene-solr/pull/1623#issuecomment-675458466
Beasting ran all night (1031 iterations) and uncovered this failure, which
reproduces with this PR but not on mainline:
```
[mkdir] Created dir:
mikemccand commented on pull request #1623:
URL: https://github.com/apache/lucene-solr/pull/1623#issuecomment-674391203
919 iterations and only this failure!
```
1> TEST: now get reader
> java.io.IOException: MockDirectoryWrapper: file "_3.cfs" is still
open: cannot
mikemccand commented on pull request #1623:
URL: https://github.com/apache/lucene-solr/pull/1623#issuecomment-674270585
256 iterations and no failures yet!
This is an automated message from the Apache Git Service.
To respond
mikemccand commented on pull request #1623:
URL: https://github.com/apache/lucene-solr/pull/1623#issuecomment-674118327
> @mikemccand nevermind I think I found the issue. I think it's ready now.
Yay! I will kick off new beasting (of all Lucene core + module tests) on my
128 core
mikemccand commented on pull request #1623:
URL: https://github.com/apache/lucene-solr/pull/1623#issuecomment-673727019
Ugh, also these test failures that I put onto the wrong PR:
https://github.com/apache/lucene-solr/pull/1743#issuecomment-673726633
mikemccand commented on pull request #1623:
URL: https://github.com/apache/lucene-solr/pull/1623#issuecomment-673725282
> I kicked off beasting of all Lucene (core + modules) tests with this
change ... no failures yet after 31 iterations.
OK, it has run 1061 iterations now of all
mikemccand commented on pull request #1623:
URL: https://github.com/apache/lucene-solr/pull/1623#issuecomment-673112665
Awesome, thanks @s1monw! I will try to have a look soon. I kicked off
beasting of all Lucene (core + modules) tests with this change ... no failures
yet after 31
mikemccand commented on pull request #1623:
URL: https://github.com/apache/lucene-solr/pull/1623#issuecomment-651167468
A different, reproducing, test failure, likely from the same cause:
```
> java.lang.AssertionError
> at
mikemccand commented on pull request #1623:
URL: https://github.com/apache/lucene-solr/pull/1623#issuecomment-651166475
Here's a fun tragic failure test that repros:
```
[junit4] 2> NOTE: reproduce with: ant test
-Dtestcase=TestIndexWriterOnVMError
mikemccand commented on pull request #1623:
URL: https://github.com/apache/lucene-solr/pull/1623#issuecomment-651156942
> So this would merge small commits on refresh?
Small segments, yes.
> I wonder if it would be more obvious to users if we call the MergeTrigger
REFRESH?
mikemccand commented on pull request #1623:
URL: https://github.com/apache/lucene-solr/pull/1623#issuecomment-650640435
Wow! This is incredibly simple! Thanks to the clean approach @s1monw
worked out for commit-on-merge, awesome! I think tests would exercise this due
to
17 matches
Mail list logo
