Mailing-List: contact issues-help@hbase.apache.org; run by ezmlm
Precedence: bulk
Date: Mon, 8 Aug 2011 18:14:27 +0000 (UTC)
From: "Ted Yu (JIRA)" <jira@apache.org>
To: issues@hbase.apache.org
Message-ID: 
 <932809881.17238.1312827267107.JavaMail.tomcat@hel.zones.apache.org>
In-Reply-To: 
 <1586267739.17174.1312825827176.JavaMail.tomcat@hel.zones.apache.org>
Subject: =?utf-8?Q?[jira]_[Commented]_(HBASE-4177)_Handling_read?=
 =?utf-8?Q?_failures_during_recovery=E2=80=8F_-_when_HMaster_call?=
 =?utf-8?Q?s_Namenode_recovery,_recovery_may_be_a_failur?=
 =?utf-8?Q?e_leading_to_read_failure_while_splitting_logs?=
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable


    [ https://issues.apache.org/jira/browse/HBASE-4177?page=3Dcom.atlassian=
.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=3D1308=
1098#comment-13081098 ]=20

Ted Yu commented on HBASE-4177:
-------------------------------

Looking at FSUtils.recoverFileLease(), we check the type of fs inside while=
 loop. This is unnecessary.

w.r.t. soft limit for the lease, we have:
{code}
          if (waitedFor > FSConstants.LEASE_SOFTLIMIT_PERIOD) {
            LOG.warn("Waited " + waitedFor + "ms for lease recovery on " + =
p +
              ":" + e.getMessage());
          }
{code}
I think we should wait for the remainder of soft limit (which is 60 seconds=
).


> Handling read failures during recovery=E2=80=8F - when HMaster calls Name=
node recovery, recovery may be a failure leading to read failure while spli=
tting logs
> -------------------------------------------------------------------------=
---------------------------------------------------------------------------=
--
>
>                 Key: HBASE-4177
>                 URL: https://issues.apache.org/jira/browse/HBASE-4177
>             Project: HBase
>          Issue Type: Bug
>          Components: master
>            Reporter: ramkrishna.s.vasudevan
>            Assignee: ramkrishna.s.vasudevan
>
> As per the mailing thread with the heading
> 'Handling read failures during recovery=E2=80=8F' we found this problem.
> As part of split Logs the HMaster calls Namenode recovery.  The recovery =
is an asynchronous process.=20
> In HDFS
> =3D=3D=3D=3D=3D=3D=3D
> Even though client is getting the updated block info from Namenode on fir=
st
> read failure, client is discarding the new info and using the old info on=
ly
> to retrieve the data from datanode. So, all the read
> retries are failing. [Method parameter reassignment - Not reflected in
> caller].=20
> In HBASE
> =3D=3D=3D=3D=3D=3D=3D
> In HMaster code we tend to wait for  1sec.  But if the recovery had some =
failure then split log may not happen and may lead to dataloss.
> So may be we need to decide upon the actual delay that needs to be introd=
uced once Hmaster calls NN recovery.

--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira