From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1760716AbaCUQhu (ORCPT ); Fri, 21 Mar 2014 12:37:50 -0400 Received: from mail.windriver.com ([147.11.1.11]:40188 "EHLO mail.windriver.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1760463AbaCUQhq (ORCPT ); Fri, 21 Mar 2014 12:37:46 -0400 Message-ID: <532C6AD8.6000406@windriver.com> Date: Fri, 21 Mar 2014 10:37:44 -0600 From: Chris Friesen User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:24.0) Gecko/20100101 Thunderbird/24.2.0 MIME-Version: 1.0 To: Subject: [bug?] can't unmount filesystem after running "exportfs -u" -- dangling refcount? Content-Type: text/plain; charset="ISO-8859-1"; format=flowed Content-Transfer-Encoding: 7bit X-Originating-IP: [147.11.118.38] Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org I'm seeing some odd behaviour on an NFS server. We're running 3.4.82 and have a drbd-replicated filesystem (active/standby) that is exported via NFS (explicitly v3, and /proc/fs/nfsd is mounted). On a server switchover we take down the NFS server IP address, run "exportfs -u" on the exports, unmount the drbd filesystem, and then demote drbd from primary to secondary. The problem we're seeing is that sometimes "exportfs -u" completes successfully but "umount" fails with an error that the filesystem is busy. When this happens, "fuser" doesn't see any userspace users, which points at the kernel thinking it has a reference open. When it's in the stuck state, manually running "exportfs -f" seems to resolve the issue so that "unmount" can run. Are there any known issue in this area? I'm not currently subscribed to the list, so please CC me on replies. Thanks, Chris