Mailing-List: contact user-help@couchdb.apache.org; run by ezmlm
Precedence: bulk
Reply-To: user@couchdb.apache.org
Received-SPF: pass (nike.apache.org: domain of rusvdw@gmail.com designates
 209.85.216.52 as permitted sender)
DomainKey-Signature: a=rsa-sha1; c=nofws;
        d=gmail.com; s=gamma;
        h=mime-version:in-reply-to:references:date:message-id:subject:from:to
         :content-type;
        b=TkiC5uv4LcIuJ+Z4yC/B9LGzl7q2GSuK2YzpWEFGDcejc6nqKapdfe5g19NaV0ySi3
         act9kFf3wcfux4wxb8acdT8FF7w4bW3irHoZDKC8NsbLkBPBPupBd5M4eJf9qiXiUtSb
         T6bhPOjEHSUxKLyMGbyGLKDtqIEgD6+pjr/tY=
MIME-Version: 1.0
In-Reply-To: <AANLkTinzCbKGt2QeHSPFOi1HvBZTRFvrtYDC_0ggM=-C@mail.gmail.com>
References: <AANLkTim0_c7KES3O-Dg5+Fb1RUJ29LN3FMqdSWiDcT3g@mail.gmail.com>
	<AANLkTimTYcneUuDtg1mdSTGp4mjzs+oxP-ZDfHG5zi=S@mail.gmail.com>
	<AANLkTinzCbKGt2QeHSPFOi1HvBZTRFvrtYDC_0ggM=-C@mail.gmail.com>
Date: Fri, 29 Oct 2010 15:43:22 +0200
Message-ID: <AANLkTimP5NMfjJE63sZ7ZTj9pscfpSKLpuyG06D_py5-@mail.gmail.com>
Subject: Re: Couch crashing every couple of days
From: Russell <rusvdw@gmail.com>
To: user@couchdb.apache.org
Content-Type: text/plain; charset=ISO-8859-1

Thanks guys, I'll definitely give the auto restart a go, but I am a
little worried about what the debug level logging is going to do to my
production servers.

On Fri, Oct 29, 2010 at 11:21 AM, Randall Leeds <randall.leeds@gmail.com> wrote:
> On Fri, Oct 29, 2010 at 02:08, Dirkjan Ochtman <dirkjan@ochtman.nl> wrote:
>> On Fri, Oct 29, 2010 at 11:05, Russell <rusvdw@gmail.com> wrote:
>>> I am well aware that a crash report with no information is useless,
>>> but I cannot find any logs that tell me what happened. So my question
>>> is:
>>>
>>> 1. Where should I be looking for information about the crash?
>>
>> You could try to turn on debug-level logs to see what happens right
>> before it goes down. Also, is there any memory pressure on the server?
>> You might also want to check your syslog.
>>
>>> 2. What is the best way to auto restart couch on Ubuntu whenever this happens?
>>
>> I think you can just use normal monitoring services (monit,
>> supervisor, that kind of thing).
>
> There's heartbeat option (-r X) which will restart the server after X
> seconds if it dies.
> If you're starting couchdb from an init script you may also find
> /etc/default/couchdb (or something equivalent in your distro) is a
> good place to set this.
>