You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@lucene.apache.org by "Cao Manh Dat (JIRA)" <ji...@apache.org> on 2018/03/01 08:47:00 UTC

[jira] [Commented] (SOLR-12012) Replicas should skip doing recovery on startup if possible

    [ https://issues.apache.org/jira/browse/SOLR-12012?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16381685#comment-16381685 ] 

Cao Manh Dat commented on SOLR-12012:
-------------------------------------

Patch for this ticket based on SOLR-12011 changes. This patch does not handle rolling updates case and the efforts to do that is a lot. I think it is much better to postpone this issue until 8.0 where we stop support rolling updates.

> Replicas should skip doing recovery on startup if possible
> ----------------------------------------------------------
>
>                 Key: SOLR-12012
>                 URL: https://issues.apache.org/jira/browse/SOLR-12012
>             Project: Solr
>          Issue Type: Improvement
>      Security Level: Public(Default Security Level. Issues are Public) 
>          Components: SolrCloud
>            Reporter: Cao Manh Dat
>            Assignee: Cao Manh Dat
>            Priority: Major
>         Attachments: SOLR-12012.patch
>
>
> Right now a replica when first loaded always does recovery (except the replica is the leader). This will lead to several problems, for example: 
> - leaderless if at the same time the replica is doing recovery, the current leader goes down.
> - the recovery process is not necessary if the replica is already in-sync with the leader
> By using term value introduced in SOLR-11702 we can skip the recovery process if replica's term equals to leader's term.



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

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