[
https://issues.apache.org/jira/browse/SOLR-15145?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17281317#comment-17281317
]
Timothy Potter edited comment on SOLR-15145 at 2/8/21, 7:43 PM:
----------------------------------------------------------------
Actually now that I think about this some more, this may also impact rolling
upgrades from pre 8.8 to an 8.8 cluster as upgrade nodes re-joining the cluster
will not persist the {{base_url}} in stored cluster state, which could cause
issues with leader election / recoveries as there will be a mix of node
versions in the cluster temporarily.
Tried this out locally and older nodes see errors like this while new nodes are
coming back up:
{code}
2021-02-08 19:42:20.107 ERROR
(OverseerStateUpdate-72081179721793543-dev-solrcloud-0.dev:80_solr-n_0000000000)
[ ] o.a.s.c.Overseer Overseer could not process the current clusterstate
state update message, skipping the message: {
"operation":"leader",
"shard":"shard1",
"collection":"sop1",
"node_name":"dev-solrcloud-2.dev:80_solr",
"core":"sop1_shard1_replica_n1",
"state":"active"} => java.lang.NullPointerException
at
org.apache.solr.common.cloud.ZkCoreNodeProps.getCoreUrl(ZkCoreNodeProps.java:53)
java.lang.NullPointerException: null
at
org.apache.solr.common.cloud.ZkCoreNodeProps.getCoreUrl(ZkCoreNodeProps.java:53)
~[?:?]
at
org.apache.solr.cloud.overseer.SliceMutator.setShardLeader(SliceMutator.java:140)
~[?:?]
at
org.apache.solr.cloud.Overseer$ClusterStateUpdater.processMessage(Overseer.java:420)
~[?:?]
at
org.apache.solr.cloud.Overseer$ClusterStateUpdater.processQueueItem(Overseer.java:309)
~[?:?]
at
org.apache.solr.cloud.Overseer$ClusterStateUpdater.run(Overseer.java:253) ~[?:?]
at java.lang.Thread.run(Unknown Source) [?:?]
{code}
was (Author: thelabdude):
Actually now that I think about this some more, this may also impact rolling
upgrades from pre 8.8 to an 8.8 cluster as upgrade nodes re-joining the cluster
will not persist the {{base_url}} in stored cluster state, which could cause
issues with leader election / recoveries as there will be a mix of node
versions in the cluster temporarily.
> Older versions of SolrJ (pre-8.8.0) hit an NPE when computing the base_url
> for core node props
> ----------------------------------------------------------------------------------------------
>
> Key: SOLR-15145
> URL: https://issues.apache.org/jira/browse/SOLR-15145
> Project: Solr
> Issue Type: Bug
> Security Level: Public(Default Security Level. Issues are Public)
> Components: SolrJ
> Affects Versions: 8.8
> Reporter: Timothy Potter
> Assignee: Timothy Potter
> Priority: Major
> Fix For: 8.8.1
>
>
> From the mailing list:
> {code}
> Caused by: java.lang.NullPointerException
> at
> deployment.uleaf.ear//org.apache.solr.common.cloud.ZkCoreNodeProps.getCoreUrl(ZkCoreNodeProps.java:53)
> at
> deployment.uleaf.ear//org.apache.solr.client.solrj.impl.BaseCloudSolrClient.lambda$sendRequest$2(BaseCloudSolrClient.java:1161)
> at java.base/java.util.ArrayList.forEach(ArrayList.java:1540)
> at
> deployment.uleaf.ear//org.apache.solr.client.solrj.impl.BaseCloudSolrClient.sendRequest(BaseCloudSolrClient.java:1159)
> at
> deployment.uleaf.ear//org.apache.solr.client.solrj.impl.BaseCloudSolrClient.requestWithRetryOnStaleState(BaseCloudSolrClient.java:934)
> ... 166 more
> {code}
> see:
> https://lists.apache.org/thread.html/r3d131030f0a7026235451f71fabdae6d6b7c2f955822c75dcad4d41f%40%3Csolr-user.lucene.apache.org%3E
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@lucene.apache.org
For additional commands, e-mail: issues-h...@lucene.apache.org
