From mboxrd@z Thu Jan 1 00:00:00 1970 From: Christoph Hellwig Subject: Re: [PATCH 1/2] scsi_transport_fc: implement 'disable_target_scan' module parameter Date: Wed, 2 Mar 2016 10:18:42 +0100 Message-ID: <20160302091842.GD1308@lst.de> References: <1456127462-2817-1-git-send-email-hare@suse.de> <1456127462-2817-2-git-send-email-hare@suse.de> <20160301131252.GA7786@lst.de> <56D6814B.9020302@suse.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Received: from verein.lst.de ([213.95.11.211]:45056 "EHLO newverein.lst.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754123AbcCBJSo (ORCPT ); Wed, 2 Mar 2016 04:18:44 -0500 Content-Disposition: inline In-Reply-To: <56D6814B.9020302@suse.de> Sender: linux-scsi-owner@vger.kernel.org List-Id: linux-scsi@vger.kernel.org To: Hannes Reinecke Cc: Christoph Hellwig , "Martin K . Petersen" , James Bottomley , Johannes Thumshirn , linux-scsi@vger.kernel.org On Wed, Mar 02, 2016 at 01:59:39PM +0800, Hannes Reinecke wrote: > Main reasons being that > a) it's most commonly on FC where you end up having thousands of LUNs; > SAS domains are of finite size and you'd need to string quite a few SAS > expanders together to achieve a similar setup. And iSCSI is ... well. > b) FC already has the required infrastructure in place where you easily > can suppress a target scan, making the patch really trivial. > > So rather than trying to come up with a generic solution and an overly > complicated patch I opted for the simpler version :-) Having a global disable just for all FC is still odd. Either you do it for all devices and always initiate a manual scan, or you need to do it on a per-host basis somehow.