Mailing-List: contact user-help@cassandra.apache.org; run by ezmlm
Precedence: bulk
Reply-To: user@cassandra.apache.org
Received-SPF: pass (nike.apache.org: domain of jbellis@gmail.com designates
 209.85.210.44 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:content-transfer-encoding;
        b=PGa3SAX322TXvJIxpDCMMnl+6JWnrCtx8wdmFLwIAoZoW2h9+SG6Z+fiJ22i4IqTUq
         OffjpOLtbnXhgHqQltL5GEaimHOvwK5Amz7VLq8x3X/pjFiA2I41Upj0dD4rp/VxyTCv
         i9tHVm2fXIP0HjYBfexjUwSrqWrTbIn1v9Fe4=
MIME-Version: 1.0
In-Reply-To: <AANLkTimnyGOYaPxsp_qzETpsW8Dab6Wag7ksY7Epw=Uu@mail.gmail.com>
References: <d24444bd-d819-4576-8f92-ca935f3882b3@me.com>
	<4CACE78C.2060807@digg.com>
	<AANLkTimPMLRHG8GmbeH35WgXzPpUxeiR=J6TUmz5tiYu@mail.gmail.com>
	<08c601cb667d$5eec4a00$1cc4de00$@com>
	<AANLkTimwFeAUmvJwXw6jQgHe_RogG2RMOpmptSvNYL-O@mail.gmail.com>
	<AANLkTimnyGOYaPxsp_qzETpsW8Dab6Wag7ksY7Epw=Uu@mail.gmail.com>
Date: Fri, 8 Oct 2010 08:06:20 -0500
Message-ID: <AANLkTimFdvAz4tShesf_4chq-kQEeEb18HbqZzHKyLXp@mail.gmail.com>
Subject: Re: Newbie Question about restarting Cassandra
From: Jonathan Ellis <jbellis@gmail.com>
To: user <user@cassandra.apache.org>
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable

On Fri, Oct 8, 2010 at 2:12 AM, Peter Schuller
<peter.schuller@infidyne.com> wrote:
>> If your C* process dies and/or is killed you should not lose data.=A0 It=
's
>> written to the commit log before the client is acked, however that entry=
 may
>> not have made it to disk yet in the case of commitlogsync=3Dperiodic.=A0=
 So, if
>> you kill the C* process you're fine.=A0 If you nicely restart the OS, yo=
u
>> should be fine (assuming your boxen/raid controllers/disks/etc do the sa=
ne
>> thing).=A0 If you nuke your OS, then see above about losing data on CL.O=
NE.
>
> Being fine in case of killing Cassandra does depend though on
> Cassandra still guaranteeing that it write():s to the fd prior to ack
> even in periodic mode. I haven't checked, but I sort of took the
> default position of assuming it did not unless otherwise specified.
>
> If it does, and if this is an intended guarantee, that's great (and
> I'll happily submit a patch to make this explicit in docs/comments).
> Can anyone confirm/deny that this is an intended guarantee?

It is not.

--=20
Jonathan Ellis
Project Chair, Apache Cassandra
co-founder of Riptano, the source for professional Cassandra support
http://riptano.com