From owner-freebsd-scsi@FreeBSD.ORG Mon Apr 5 20:34:54 2010 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 2FE1E1065677 for ; Mon, 5 Apr 2010 20:34:54 +0000 (UTC) (envelope-from mj@feral.com) Received: from ns1.feral.com (ns1.feral.com [192.67.166.1]) by mx1.freebsd.org (Postfix) with ESMTP id 076A88FC0A for ; Mon, 5 Apr 2010 20:34:53 +0000 (UTC) Received: from [192.168.221.2] (remotevpn [192.168.221.2]) by ns1.feral.com (8.14.3/8.14.3) with ESMTP id o35KYqbW013731 (version=TLSv1/SSLv3 cipher=DHE-RSA-CAMELLIA256-SHA bits=256 verify=NO) for ; Mon, 5 Apr 2010 13:34:53 -0700 (PDT) (envelope-from mj@feral.com) Message-ID: <4BBA496C.4040908@feral.com> Date: Mon, 05 Apr 2010 13:34:52 -0700 From: Matthew Jacob Organization: Feral Software User-Agent: Mozilla/5.0 (X11; U; Linux x86_64; en-US; rv:1.9.1.9) Gecko/20100330 Fedora/3.0.4-1.fc11 Thunderbird/3.0.4 MIME-Version: 1.0 To: freebsd-scsi@freebsd.org References: <1269900407.00235175.1269888603@10.7.7.3> <4BB98522.6000704@FreeBSD.org> In-Reply-To: <4BB98522.6000704@FreeBSD.org> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit X-Greylist: Sender DNS name whitelisted, not delayed by milter-greylist-4.2.3 (ns1.feral.com [192.168.221.1]); Mon, 05 Apr 2010 13:34:53 -0700 (PDT) Subject: Re: adding a "retry command after a delay" error X-BeenThere: freebsd-scsi@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list Reply-To: mj@feral.com List-Id: SCSI subsystem List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 05 Apr 2010 20:34:54 -0000 On 04/04/2010 11:37 PM, Alexander Motin wrote: > Matthew Jacob wrote: > >> This is something I whipped up for dealing with some >> active/active-after-failover-time systems. I wanted to have a general >> facility to say of an ASC/ASCQ- retry the command, but after a period of >> delay. >> >> Wonder if anyone had comments? >> http://people.freebsd.org/~mjacob/delay_diffs.txt >> > I am not SCSI expert yet, but it seems OK for me. But I see two related > questions there: delay time and number of retries allowed. > > Most of peripheral drivers now AFAIR ask to do only few (2-5) retries, > so if delay will be short - it may give up too early, but long delay > will probably reduce performance. > > I've been doing some rethinking about this. It really is the devq that I want to have stopped and retried with a timeout- not the entire sim. Your point about periph drivers is well taken, and I thought about just capturing this one in da, but it turns out that both sg && pass have the same issue, hence the notion of generalizing it. The sim change is working for the Panasas guys right now, but that's just a bandaid. I'll think some more about this one. Thanks for the comments!