You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@lucene.apache.org by "Alan Woodward (JIRA)" <ji...@apache.org> on 2014/06/30 23:05:24 UTC

[jira] [Updated] (SOLR-6217) Load cores asynchronously on init

     [ https://issues.apache.org/jira/browse/SOLR-6217?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Alan Woodward updated SOLR-6217:
--------------------------------

    Attachment: SOLR-6217.patch

Patch making all core init happen in the background.  CoreContainer.load() creates a threadpool as before, but submits all core load processes without waiting for their responses, shuts down the threadpool and returns.

> Load cores asynchronously on init
> ---------------------------------
>
>                 Key: SOLR-6217
>                 URL: https://issues.apache.org/jira/browse/SOLR-6217
>             Project: Solr
>          Issue Type: Improvement
>            Reporter: Alan Woodward
>            Assignee: Alan Woodward
>            Priority: Minor
>         Attachments: SOLR-6217.patch
>
>
> At the moment, we load cores in a threadpool in CoreContainer.init(), but then wait for all cores to load before we register them in ZK and return.  This has two drawbacks:
> 1) SolrDispatchFilter isn't fully initialised until all the cores are loaded, which means that we can still have cores registered in ZK that can't accept HTTP connections
> 2) One slow-loading core will slow down all cores in the node.



--
This message was sent by Atlassian JIRA
(v6.2#6252)

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