From mboxrd@z Thu Jan 1 00:00:00 1970 From: Alex Elder Subject: Re: crash in rbd_img_request_create Date: Tue, 13 May 2014 12:18:18 -0500 Message-ID: <537253DA.4030903@ieee.org> References: <536EEA5C.70503@ieee.org> <53704F70.3090800@ieee.org> <537211AD.9070209@ieee.org> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 7bit Return-path: Received: from mail-ig0-f181.google.com ([209.85.213.181]:59003 "EHLO mail-ig0-f181.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753555AbaEMRSI (ORCPT ); Tue, 13 May 2014 13:18:08 -0400 Received: by mail-ig0-f181.google.com with SMTP id h3so672945igd.14 for ; Tue, 13 May 2014 10:18:07 -0700 (PDT) In-Reply-To: Sender: ceph-devel-owner@vger.kernel.org List-ID: To: Hannes Landeholm , Ilya Dryomov , Ceph Development Cc: Thorwald Lundqvist On 05/13/2014 12:17 PM, Hannes Landeholm wrote: > On Tue, May 13, 2014 at 2:35 PM, Alex Elder wrote: >> On 05/12/2014 12:28 PM, Hannes Landeholm wrote: >>> We where experimenting to use ZFS as a backend for >>> CEPH just to see what the performance and storage saving >>> characteristics was like but we're gonna completely switch to ext4 now >>> and see if the problem goes away. >> >> This is very good to know. Please report back what you find. > > We completely cleansed the server in question from any trace of ZFS > and ran the same test again with ext4 instead for the ceph backend > with the same conditions. Everything worked fine. The machine has also > stopped randomly locking up. We will not touch ZFS again. Maybe in a > distant future where it actually integrates well with Linux. That's good news for rbd... Thanks a lot. -Alex > Thank you for your time, > Hannes >