[jira] [Commented] (HBASE-14146) Once replication sees an error it slows down forever

["Matteo Bertozzi (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/HBASE-14146?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14637714#comment-14637714 ] 

Matteo Bertozzi commented on HBASE-14146:
-----------------------------------------

is that 0 instead of 1 doing what you expect?
I see that we have stuff like:
{code}
if (sleepForRetries("ReplicationEndpoint threw exception", sleepMultiplier)) {
   sleepMultiplier++;
}
protected boolean sleepForRetries(String msg, int sleepMultiplier) {
  ...
  Thread.sleep(this.sleepForRetries * sleepMultiplier);
}
{code}
is the first sleep of 0 expected?

> Once replication sees an error it slows down forever
> ----------------------------------------------------
>
>                 Key: HBASE-14146
>                 URL: https://issues.apache.org/jira/browse/HBASE-14146
>             Project: HBase
>          Issue Type: Bug
>    Affects Versions: 2.0.0, 1.2.0
>            Reporter: Elliott Clark
>            Assignee: Elliott Clark
>         Attachments: HBASE-14146.patch
>
>
> sleepMultiplier inside of HBaseInterClusterReplicationEndpoint and ReplicationSource never gets reset to zero.



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