Mailing-List: contact user-help@hadoop.apache.org; run by ezmlm
Precedence: bulk
Reply-To: user@hadoop.apache.org
Received-SPF: pass (nike.apache.org: domain of dechouxb@gmail.com designates
 209.85.217.176 as permitted sender)
MIME-Version: 1.0
In-Reply-To: <51472A17.2090001@getjar.com>
References: <51472A17.2090001@getjar.com>
Date: Mon, 18 Mar 2013 16:28:11 +0100
Message-ID: 
 <CAO6W-2cAhGwrVfT0+Upg7G993r7A8tE+AhfMpuiwfM4QeUaBUA@mail.gmail.com>
Subject: Re: namenode directory failure question
From: Bertrand Dechoux <dechouxb@gmail.com>
To: user@hadoop.apache.org
Content-Type: multipart/alternative; boundary=bcaec554dbc8fcaa7104d834a282

--bcaec554dbc8fcaa7104d834a282
Content-Type: text/plain; charset=ISO-8859-1

You may want to check this JIRA:
https://issues.apache.org/jira/browse/HADOOP-4885

It won't help you right know but it could allow you next time to avoid
restarting.

Regards

Bertrand

On Mon, Mar 18, 2013 at 3:52 PM, Brennon Church <brennon@getjar.com> wrote:

> Hello all,
>
> We have our dfs.name.dir configured to write to two local and one NFS
> directories.  The NFS server in question had to be restarted a couple days
> back and that copy of the namenode data fell behind as a result.  As I
> understand it, restarting hadoop will take the most recent copy of the
> namenode data, in this case one of the two local copies, and write that to
> all three locations going forward.  So that solves the problem.
>
> My question is this, is there a way to get the NFS copy of the data back
> in sync without having to shut down and restart the namenode? I'd prefer to
> not take an outage if I can help it.
>
> Thanks.
>
> --Brennon
>
>

--bcaec554dbc8fcaa7104d834a282
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable

You may want to check this JIRA: <a href=3D"https://issues.apache.org/jira/=
browse/HADOOP-4885">https://issues.apache.org/jira/browse/HADOOP-4885</a><b=
r><br>It won&#39;t help you right know but it could allow you next time to =
avoid restarting.<br>
<br>Regards<br><br>Bertrand<br><br><div class=3D"gmail_quote">On Mon, Mar 1=
8, 2013 at 3:52 PM, Brennon Church <span dir=3D"ltr">&lt;<a href=3D"mailto:=
brennon@getjar.com" target=3D"_blank">brennon@getjar.com</a>&gt;</span> wro=
te:<br>
<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1p=
x #ccc solid;padding-left:1ex">Hello all,<br>
<br>
We have our dfs.name.dir configured to write to two local and one NFS direc=
tories. =A0The NFS server in question had to be restarted a couple days bac=
k and that copy of the namenode data fell behind as a result. =A0As I under=
stand it, restarting hadoop will take the most recent copy of the namenode =
data, in this case one of the two local copies, and write that to all three=
 locations going forward. =A0So that solves the problem.<br>

<br>
My question is this, is there a way to get the NFS copy of the data back in=
 sync without having to shut down and restart the namenode? I&#39;d prefer =
to not take an outage if I can help it.<br>
<br>
Thanks.<span class=3D"HOEnZb"><font color=3D"#888888"><br>
<br>
--Brennon<br>
<br>
</font></span></blockquote></div><br><br>

--bcaec554dbc8fcaa7104d834a282--