You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@geode.apache.org by "Dan Smith (JIRA)" <ji...@apache.org> on 2016/03/24 19:28:25 UTC

[jira] [Commented] (GEODE-1117) Gateway sender manual-start=true causes region data not to be recovered from disk until gateway sender is started

    [ https://issues.apache.org/jira/browse/GEODE-1117?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15210715#comment-15210715 ] 

Dan Smith commented on GEODE-1117:
----------------------------------

The issue is that the gateway sender does not actually create the queue region until start is called.

Persistent recovery does not happen until all colocated regions are created - this includes the gateway queue region if it is persistent. This is to avoid moving data in some of the colocated regions before all of the colocated regions have actually been recovered - otherwise we would break the colocation and bad things would happen.

We should look into why the gateway sender doesn't create the queue region until start is called and see if it can create the queue region earlier.

> Gateway sender manual-start=true causes region data not to be recovered from disk until gateway sender is started
> -----------------------------------------------------------------------------------------------------------------
>
>                 Key: GEODE-1117
>                 URL: https://issues.apache.org/jira/browse/GEODE-1117
>             Project: Geode
>          Issue Type: Bug
>          Components: wan
>            Reporter: William Markito Oliveira
>
> Gateway sender manual-start=true causes region data not to be recovered from disk until gateway sender is started which results in a cluster with no data.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)