You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@hbase.apache.org by "Jimmy Xiang (JIRA)" <ji...@apache.org> on 2014/04/07 20:45:17 UTC

[jira] [Resolved] (HBASE-10816) CatalogTracker abortable usage

     [ https://issues.apache.org/jira/browse/HBASE-10816?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Jimmy Xiang resolved HBASE-10816.
---------------------------------

    Resolution: Won't Fix

I looked into it, and prefer not to do anything for now.

> CatalogTracker abortable usage
> ------------------------------
>
>                 Key: HBASE-10816
>                 URL: https://issues.apache.org/jira/browse/HBASE-10816
>             Project: HBase
>          Issue Type: Bug
>            Reporter: Jimmy Xiang
>            Assignee: Jimmy Xiang
>            Priority: Minor
>
> In reviewing patch for HBASE-10569, Stack pointed out some existing issue with CatalogTracker. I looked into it and I think the abortable usage can be improved.
> * If ZK is null, when a new one is created, the abortable could be null. We need consider this.
> * The throwableAborter is to abort the process in case some ZK exception in MetaRegionTracker. In case the tracker is in a server, we don't need to do this, we can use the server as the abortable. In case the tracker is in a client, we can just abort the connection. Right?



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