[
https://issues.apache.org/jira/browse/SOLR-12986?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Hoss Man reassigned SOLR-12986:
-------------------------------
Assignee: Cao Manh Dat
Dat, can you please take a look at this?
I suspect that maybe the issue has to do with the replica types being used?
In the log from the failing seed that i attached (and other failing seds i've
seen) it's using NRT replicas...
{noformat}
[junit4] 2> 35846 INFO (qtp1233081730-94) [n:127.0.0.1:51155_ ]
o.a.s.h.a.CollectionsHandler Invoked Collection Action :create with params
pullReplicas=0&maxShardsPerNode=1&collection.configName=conf1&nrtReplicas=2&name=c8n_1x2&action=CREATE&numShards=1&tlogReplicas=0&stateFormat=1&wt=javabin&version=2
and sendToOCPQueue=true
{noformat}
... but in at least 2 other seeds i've tried where the test passes it's using
tlog replicas ...
{noformat}
ant test -Dtestcase=HttpPartitionTest -Dtests.method=test
-Dtests.seed=D6EF839EC5466577 -Dtests.slow=true -Dtests.badapples=true
-Dtests.locale=ar-OM -Dtests.timezone=America/Los_Angeles -Dtests.asserts=true
-Dtests.file.encoding=UTF-8
...
[junit4] 2> 34178 INFO (qtp2003220851-136) [n:127.0.0.1:34609_fv ]
o.a.s.h.a.CollectionsHandler Invoked Collection Action :create with params
pullReplicas=0&maxShardsPerNode=1&collection.configName=conf1&nrtReplicas=0&name=c8n_1x2&action=CREATE&numShards=1&tlogReplicas=2&stateFormat=1&wt=javabin&version=2
and sendToOCPQueue=true
{noformat}
> HttpPartitionTest reliably failing seed
> ---------------------------------------
>
> Key: SOLR-12986
> URL: https://issues.apache.org/jira/browse/SOLR-12986
> Project: Solr
> Issue Type: Bug
> Security Level: Public(Default Security Level. Issues are Public)
> Reporter: Hoss Man
> Assignee: Cao Manh Dat
> Priority: Major
> Attachments: HttpPartitionTest.fail.log.txt
>
>
> I've found a reliably failing seed for HttpPartitionTest when run against
> master @ e81dd4e870d2a9b27e1f4366e92daa6dba054da8 ...
> {noformat}
> [junit4] 2> NOTE: reproduce with: ant test -Dtestcase=HttpPartitionTest
> -Dtests.method=test -Dtests.seed=E3596048AE430BB6 -Dtests.slow=true
> -Dtests.badapples=true -Dtests.locale=ar-OM
> -Dtests.timezone=America/Los_Angeles -Dtests.asserts=true
> -Dtests.file.encoding=UTF-8
> [junit4] FAILURE 56.2s | HttpPartitionTest.test <<<
> [junit4] > Throwable #1: java.lang.AssertionError: max version bucket
> seed not updated after recovery!
> [junit4] > at
> __randomizedtesting.SeedInfo.seed([E3596048AE430BB6:6B0D5F9200BF664E]:0)
> [junit4] > at
> org.apache.solr.cloud.HttpPartitionTest.testRf2(HttpPartitionTest.java:290)
> [junit4] > at
> org.apache.solr.cloud.HttpPartitionTest.test(HttpPartitionTest.java:131)
> {noformat}
> I'll attach some logs, but at first glance it appears that everything is
> working properly, and PeerSync hapens from the leader before the replica
> registers itself as active, but then it stil reports an old "max version"
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
