From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1755840AbdKCQ3P (ORCPT ); Fri, 3 Nov 2017 12:29:15 -0400 Received: from userp1040.oracle.com ([156.151.31.81]:45753 "EHLO userp1040.oracle.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752321AbdKCQ3N (ORCPT ); Fri, 3 Nov 2017 12:29:13 -0400 To: Cathy Avery Cc: kys@microsoft.com, hch@infradead.org, haiyangz@microsoft.com, jejb@linux.vnet.ibm.com, martin.petersen@oracle.com, sthemmin@microsoft.com, dan.carpenter@oracle.com, devel@linuxdriverproject.org, linux-kernel@vger.kernel.org, linux-scsi@vger.kernel.org, longli@microsoft.com, tj@kernel.org Subject: Re: [PATCH V3] scsi: storvsc: Allow only one remove lun work item to be issued per lun From: "Martin K. Petersen" Organization: Oracle Corporation References: <1509454326-11118-1-git-send-email-cavery@redhat.com> Date: Fri, 03 Nov 2017 12:28:30 -0400 In-Reply-To: <1509454326-11118-1-git-send-email-cavery@redhat.com> (Cathy Avery's message of "Tue, 31 Oct 2017 08:52:06 -0400") Message-ID: User-Agent: Gnus/5.13 (Gnus v5.13) Emacs/25.3 (gnu/linux) MIME-Version: 1.0 Content-Type: text/plain X-Source-IP: userv0021.oracle.com [156.151.31.71] Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Cathy, > When running multipath on a VM if all available paths go down the > driver can schedule large amounts of storvsc_remove_lun work items to > the same lun. In response to the failing paths typically storvsc > responds by taking host->scan_mutex and issuing a TUR per lun. If > there has been heavy IO to the failed device all the failed IOs are > returned from the host. A remove lun work item is issued per failed > IO. If the outstanding TURs have not been completed in a timely manner > the scan_mutex is never released or released too late. Consequently > the many remove lun work items are not completed as scsi_remove_device > also tries to take host->scan_mutex. This results in dragging the VM > down and sometimes completely. > > This patch only allows one remove lun to be issued to a particular lun > while it is an instantiated member of the scsi stack. Applied to 4.15/scsi-queue. Next time the change log needs to go after a "---" delimiter. Thank you! > Changes since v1: > Use single threaded workqueue to serialize work in > storvsc_handle_error [Christoph Hellwig] > > Changes since v2: > Replaced create_singlethread_workqueue with > alloc_ordered_workqueue [Christoph Hellwig] > > Added reviewed by's. -- Martin K. Petersen Oracle Linux Engineering