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 akelpe@concurrentinc.com
 designates 209.85.220.178 as permitted sender)
MIME-Version: 1.0
In-Reply-To: 
 <CAFAd71Vp8JE06EZQkX289JL7wUpedDb1JndQF5Wr-fbM9am0MQ@mail.gmail.com>
References: 
 <CAFAd71Vp8JE06EZQkX289JL7wUpedDb1JndQF5Wr-fbM9am0MQ@mail.gmail.com>
Date: Wed, 17 Sep 2014 13:03:30 +0200
Message-ID: 
 <CAA2tiYLh0RTWE993iX3kC787trTdWppbWq7Xq2UNtJeYp4JSyg@mail.gmail.com>
Subject: Re: hadoop cluster crash problem
From: Andre Kelpe <akelpe@concurrentinc.com>
To: user@hadoop.apache.org
Content-Type: multipart/alternative; boundary=20cf307f31667e66ae050340d1d3

--20cf307f31667e66ae050340d1d3
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

virtualbox is known for causing instabilities in the host-kernel (or at
least, it used to). You might be better off asking for support there:
https://www.virtualbox.org/wiki/Bugtracker

- Andr=C3=A9

On Wed, Sep 17, 2014 at 4:25 AM, Li Li <fancyerii@gmail.com> wrote:

> hi all,
>     I know it's not a problem related to hadoop but administrator can
> not find any clues.
>     I have a machine with 24 core and 64GB memory with ubuntu 12.04
> LTS. we use virtual box to create 4 virtual machine. Each vm has 10GB
> memory and 6 core.
>     I have setup a small hadoop 1.2.1 cluster with one
> jobtracker/namenode and 3 tasktracker/datanode. Each tasktrack has 4
> mapper slots and 4 reducers slot.
>     But it always crashs(the host machine crash, not vm crash).
> Sometimes it crashes for the first map-reduce job. Sometimes it can
> run a few jobs.
>     is there any clues? I have checked the sys log and can find any
> thing useful. Using monitor system, The cpu and io is not abnormal.
> The only abnormal phenomenon is context switch is high. about 40k.
>


--=20
Andr=C3=A9 Kelpe
andre@concurrentinc.com
http://concurrentinc.com

--20cf307f31667e66ae050340d1d3
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><div>virtualbox is known for causing instabilities in the =
host-kernel (or at least, it used to). You might be better off asking for s=
upport there: <a href=3D"https://www.virtualbox.org/wiki/Bugtracker">https:=
//www.virtualbox.org/wiki/Bugtracker</a><br><br></div>- Andr=C3=A9<br></div=
><div class=3D"gmail_extra"><br><div class=3D"gmail_quote">On Wed, Sep 17, =
2014 at 4:25 AM, Li Li <span dir=3D"ltr">&lt;<a href=3D"mailto:fancyerii@gm=
ail.com" target=3D"_blank">fancyerii@gmail.com</a>&gt;</span> wrote:<br><bl=
ockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1px #=
ccc solid;padding-left:1ex">hi all,<br>
=C2=A0 =C2=A0 I know it&#39;s not a problem related to hadoop but administr=
ator can<br>
not find any clues.<br>
=C2=A0 =C2=A0 I have a machine with 24 core and 64GB memory with ubuntu 12.=
04<br>
LTS. we use virtual box to create 4 virtual machine. Each vm has 10GB<br>
memory and 6 core.<br>
=C2=A0 =C2=A0 I have setup a small hadoop 1.2.1 cluster with one<br>
jobtracker/namenode and 3 tasktracker/datanode. Each tasktrack has 4<br>
mapper slots and 4 reducers slot.<br>
=C2=A0 =C2=A0 But it always crashs(the host machine crash, not vm crash).<b=
r>
Sometimes it crashes for the first map-reduce job. Sometimes it can<br>
run a few jobs.<br>
=C2=A0 =C2=A0 is there any clues? I have checked the sys log and can find a=
ny<br>
thing useful. Using monitor system, The cpu and io is not abnormal.<br>
The only abnormal phenomenon is context switch is high. about 40k.<br>
</blockquote></div><br><br clear=3D"all"><br>-- <br>Andr=C3=A9 Kelpe<br><a =
href=3D"mailto:andre@concurrentinc.com" target=3D"_blank">andre@concurrenti=
nc.com</a><br><a href=3D"http://concurrentinc.com" target=3D"_blank">http:/=
/concurrentinc.com</a>
</div>

--20cf307f31667e66ae050340d1d3--