Mailing-List: contact common-user-help@hadoop.apache.org; run by ezmlm
Precedence: bulk
Reply-To: common-user@hadoop.apache.org
Received-SPF: pass (athena.apache.org: domain of qwertymaniac@gmail.com
 designates 209.85.161.48 as permitted sender)
DomainKey-Signature: a=rsa-sha1; c=nofws;
        d=gmail.com; s=gamma;
        h=mime-version:in-reply-to:references:from:date:message-id:subject:to
         :content-type:content-transfer-encoding;
        b=NSSXvMwiXTP9d/FCtercBplLlDtP4aevY2s8cDbb2ENmiXCJqYrAc4Orq6/ogio+Um
         dkIBLhTTtjP1d8zhPabbI2RP7G9BR43g4dNAmLiC1X06CK4pUpfaVcKzlLYtJqjoAMVR
         y4lAss19whvMHSBUfIVwnaWDEJMOUSO4J7t6I=
MIME-Version: 1.0
In-Reply-To: <58142C191905468D88BB56CB8AC060AC@RINGRING>
References: <58142C191905468D88BB56CB8AC060AC@RINGRING>
From: Harsh J <qwertymaniac@gmail.com>
Date: Fri, 8 Apr 2011 23:28:01 +0530
Message-ID: <BANLkTinpwAgy2q6evFgV2E-NAOWkfEPJKA@mail.gmail.com>
Subject: Re: Re:HDFS start-up with safe mode?
To: common-user@hadoop.apache.org
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable

Hello,

I'm not quite clear why you'd want to disable a consistency check such
as the safemode feature. It is to guarantee that your DFS is to be
made ready only after it has sufficient blocks reported to start
handling your dfs requests. If your NN ever goes into safemode later,
it is vital that you take a look at logs and fsck reports to determine
what's gone wrong.

On Fri, Apr 8, 2011 at 3:06 PM, springring <springring@126.com> wrote:
> I modify the value of "dfs.safemode.threshold.pct" to zero, now everythin=
g is ok.
> log file as below
> But there are still three questions
>
> =A01.. Can I regain percentage of blocks that should satisfy the minimal =
replication requirement
> =A0 =A0 =A0 =A0 =A0to 99.9%? =A0hadoop balancer? For I feel it will be mo=
re safe.

The safemode is to guarantee that. That is why it is called the 'safe'
mode. Not sure what you mean by the balancer thing.

In production one never restarts the NameNode frequently, so I s'pose
this is just to get rid of some development hassles?

You may want to additionally lower the leave-safemode extension period
from 30s to 0s to get rid of the check entirely, anyway.

--=20
Harsh J
http://harshj.com