From owner-freebsd-scsi@FreeBSD.ORG Mon Jul 14 16:31:52 2008 Return-Path: Delivered-To: freebsd-scsi@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id 8FBD0106567A for ; Mon, 14 Jul 2008 16:31:52 +0000 (UTC) (envelope-from westr@connection.ca) Received: from nc-tor-mail1.connection.ca (nc-tor-mail1.connection.ca [205.207.122.26]) by mx1.freebsd.org (Postfix) with ESMTP id 819278FC20 for ; Mon, 14 Jul 2008 16:31:52 +0000 (UTC) (envelope-from westr@connection.ca) Received: from localhost (external.tor.connection.ca [216.234.38.18]) by nc-tor-mail1.connection.ca (Postfix) with ESMTP id C18A544B426; Mon, 14 Jul 2008 12:31:51 -0400 (EDT) Date: Mon, 14 Jul 2008 12:31:28 -0400 From: Ross Organization: Network Connection X-Priority: 3 (Normal) Message-ID: <1511139894.20080714123128@connection.ca> To: Sean Bruno In-Reply-To: <487B7D14.8020505@miralink.com> References: <1557532186.20080710145356@connection.ca> <48767C9B.3060804@miralink.com> <477336220.20080710172453@connection.ca> <487B7D14.8020505@miralink.com> MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Cc: freebsd-scsi@freebsd.org Subject: Re[2]: isp - error/scsi message X-BeenThere: freebsd-scsi@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list Reply-To: Ross List-Id: SCSI subsystem List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 14 Jul 2008 16:31:52 -0000 SB> Ross wrote: >>>> -= >>>> kernel: (da0:isp0:0:0:1): WRITE(10). CDB: 2a 0 1 10 eb bf 0 0 20 0 >>>> kernel: (da0:isp0:0:0:1): CAM Status: SCSI Status Error >>>> kernel: (da0:isp0:0:0:1): SCSI Status: Check Condition >>>> kernel: (da0:isp0:0:0:1): UNIT ATTENTION asc:2a,6 >>>> kernel: (da0:isp0:0:0:1): Reserved ASC/ASCQ pair >>>> kernel: (da0:isp0:0:0:1): Retrying Command (per Sense Data) >>>> -= SB> Is it possible that something in one of the PATH's is failing? Might be possible, but I don't think so - everything is heavily monitored (HP c7000 blade w/brocade switch into dual controller EVA). We're also using the gmultipath geom package to handle path failure, and it hasn't triggered at all. (it fails over nicely in testing at least). SB> Moreover, is the LUN failing due to this message, or is this just a SB> message that you are seeing now and again? It's just an occasional message and doesn't seem to cause any performance degradation - I'm wondering if it's the dual active/active controller on the EVA trying to force some load balancing or something. Basically I'm just trying to find out what the actual message means or in what circumstances it would be generated, so I can do some more hunting. Cheers, Ross. -- Ross