[ 
https://issues.apache.org/jira/browse/SOLR-12187?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16434439#comment-16434439
 ] 

Tomás Fernández Löbbe commented on SOLR-12187:
----------------------------------------------

Good finding. With this patch, a core would be unloaded in two ways:
1) by an UNLOAD command issued from the {{DeleteReplicaCmd}}
2) by a change to the clusterstate that triggers the state watcher, which 
unloads the core. 
Am I right? Do you think we need to keep #1?
Also, a thought on the test, should it also cover the legacyCloud=true case?

> Replica should watch clusterstate and unload itself if its entry is removed
> ---------------------------------------------------------------------------
>
>                 Key: SOLR-12187
>                 URL: https://issues.apache.org/jira/browse/SOLR-12187
>             Project: Solr
>          Issue Type: Bug
>      Security Level: Public(Default Security Level. Issues are Public) 
>            Reporter: Cao Manh Dat
>            Assignee: Cao Manh Dat
>            Priority: Major
>         Attachments: SOLR-12187.patch, SOLR-12187.patch, SOLR-12187.patch, 
> SOLR-12187.patch
>
>
> With the introduction of autoscaling framework, we have seen an increase in 
> the number of issues related to the race condition between delete a replica 
> and other stuff.
> Case 1: DeleteReplicaCmd failed to send UNLOAD request to a replica, 
> therefore, forcefully remove its entry from clusterstate, but the replica 
> still function normally and be able to become a leader -> SOLR-12176
> Case 2:
>  * DeleteReplicaCmd enqueue a DELETECOREOP (without sending a request to 
> replica because the node is not live)
>  * The node start and the replica get loaded
>  * DELETECOREOP has not processed hence the replica still present in 
> clusterstate --> pass checkStateInZk
>  * DELETECOREOP is executed, DeleteReplicaCmd finished
>  ** result 1: the replica start recovering, finish it and publish itself as 
> ACTIVE --> state of the replica is ACTIVE
>  ** result 2: the replica throw an exception (probably: NPE) 
> --> state of the replica is DOWN, not join leader election



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org
For additional commands, e-mail: dev-h...@lucene.apache.org

Reply via email to