Mailing-List: contact issues-help@hbase.apache.org; run by ezmlm
Precedence: bulk
Date: Sat, 27 Dec 2014 17:23:13 +0000 (UTC)
From: "stack (JIRA)" <jira@apache.org>
To: issues@hbase.apache.org
Message-ID: <JIRA.12763326.1419287327000.105028.1419700993969@Atlassian.JIRA>
In-Reply-To: <JIRA.12763326.1419287327000@Atlassian.JIRA>
References: <JIRA.12763326.1419287327000@Atlassian.JIRA>
 <JIRA.12763326.1419287327596@arcas>
Subject: [jira] [Commented] (HBASE-12746) [1.0.0RC0] Distributed Log Replay
 is on (HBASE-12577 was insufficient)
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit


    [ https://issues.apache.org/jira/browse/HBASE-12746?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14259427#comment-14259427 ] 

stack commented on HBASE-12746:
-------------------------------

This is an important change:

{code}
240	          FileStatus[] curLogFiles = FSUtils.listStatus(this.fs, status.getPath(), null);
241	          if (curLogFiles == null || curLogFiles.length == 0) {
242	            // Empty log folder. No recovery needed
243	            continue;
244	          }
{code}

I've seen a bunch of cases where this has Master thinking it is joining a running cluster when it is actually a fresh start. This should go back to 0.98 too, would you agree [~jeffreyz]?

Otherwise, +1 on this patch.  Let me commit.

> [1.0.0RC0] Distributed Log Replay is on (HBASE-12577 was insufficient)
> ----------------------------------------------------------------------
>
>                 Key: HBASE-12746
>                 URL: https://issues.apache.org/jira/browse/HBASE-12746
>             Project: HBase
>          Issue Type: Bug
>          Components: wal
>    Affects Versions: 1.0.0
>            Reporter: stack
>            Assignee: stack
>            Priority: Critical
>             Fix For: 1.0.0
>
>         Attachments: 12746-v2.patch, 12746.txt, 12746.txt
>
>
> Testing the 1.0.0RC0 candidate, I noticed DLR was on (because I was bumping into HBASE-12743)  I thought it my environment but apparently not.
> If I add this to HMaster....
> diff --git a/hbase-server/src/main/java/org/apache/hadoop/hbase/master/HMaster.java b/hbase-server/src/main/java/org/apache/hadoop/hbase/master/HMaster.java
> index a85c2e7..d745f94 100644
> --- a/hbase-server/src/main/java/org/apache/hadoop/hbase/master/HMaster.java
> +++ b/hbase-server/src/main/java/org/apache/hadoop/hbase/master/HMaster.java
> @@ -416,6 +416,10 @@ public class HMaster extends HRegionServer implements MasterServices, Server {
>        throw new IOException("Failed to start redirecting jetty server", e);
>      }
>      masterInfoPort = connector.getPort();
> +     boolean dlr =
> +     conf.getBoolean(HConstants.DISTRIBUTED_LOG_REPLAY_KEY,
> +      HConstants.DEFAULT_DISTRIBUTED_LOG_REPLAY_CONFIG);
> +      LOG.info("Distributed log replay=" + dlr);
>    }
> It says DLR is on.  HBASE-12577 was not enough it seems.  The hbase-default.xml still has DLR as true.


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