From mboxrd@z Thu Jan 1 00:00:00 1970 From: Thomas Glanzmann Subject: Re: [PATCH 0/3] target: Fixes for COMPARE_AND_WRITE backend I/O failure cases Date: Thu, 3 Oct 2013 13:38:59 +0200 Message-ID: <20131003113859.GB541@glanzmann.de> References: <1380672796-437-1-git-send-email-nab@daterainc.com> <20131002043157.GA3057@glanzmann.de> <1380691034.19256.36.camel@haakon3.risingtidesystems.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Content-Disposition: inline In-Reply-To: <1380691034.19256.36.camel@haakon3.risingtidesystems.com> Sender: target-devel-owner@vger.kernel.org To: "Nicholas A. Bellinger" Cc: "Nicholas A. Bellinger" , target-devel , linux-scsi List-Id: linux-scsi@vger.kernel.org Hallo Nab, > Please let me know if you encounter any issues, and a patch to address > this bit should be along in the next 24-48 hours. I just did a little bit of stress testing: - Rescan from 12 Initiators at the same time - PASS - Deployed 24 VMs over 12 Initiators at the same time 200 MB/s frontend traffic - PASS - Booted 24 VMs on 12 ESX servers at the same time - 120 MB/s frontend traffic - PASS - Did 24 svMotion at the same time - 400 MB/s backend traffic - 10 MB/s frontend traffic using XCOPY. - PASS - Unloaded the target in production, ran 'grep se_tmr /proc/slabinfo' and loaded it again. And made sure that the VMs continued to run. - PASS It looks pretty stable to me. I'll use it next week in a class and report back if I had any issues whatsoever. This was with: v3.12-rc3-4-g8a77fe9 Cheers, Thomas