lucene-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Yonik Seeley (Commented) (JIRA)" <>
Subject [jira] [Commented] (SOLR-3167) Allow running embedded zookeeper 1 for 1 dynamically with solr nodes
Date Mon, 27 Feb 2012 15:02:49 GMT


Yonik Seeley commented on SOLR-3167:

bq. Have you considered cluster auto-discover so a new node can be bootsrapped with no ZK
host/port info at all?

Yeah, that's really part of this issue.  Once we can dynamically change the ZK list, we need
to pair that with a discovery mechanism so one can just start up a node (with possibly an
optional cluster ID) and solr can find the existing ZK quorum, create a new internal ZK server
if needed, add it to the ZK quorum, etc.

Or maybe you were taking about bootstrapping w/ no ZK host/port info when you've configured
an external ZK quorum?  Haven't thought about that (or if it would make sense or not).
> Allow running embedded zookeeper 1 for 1 dynamically with solr nodes
> --------------------------------------------------------------------
>                 Key: SOLR-3167
>                 URL:
>             Project: Solr
>          Issue Type: Improvement
>            Reporter: Mark Miller
>            Assignee: Mark Miller
>             Fix For: 4.0
> Right now you have to decide which nodes run zookeeper up front - each node must know
the list of all the servers in the ensemble. Growing or shrinking the list of nodes requires
a rolling restart. (Add zk.updateServerList(newServerList)
might be able to help us here. Perhaps the over seer could make a call to each replica when
the list changes and use the update server list call.

This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators:!default.jspa
For more information on JIRA, see:


To unsubscribe, e-mail:
For additional commands, e-mail:

View raw message