Return-Path: linux-nfs-owner@vger.kernel.org Received: from peace.netnation.com ([204.174.223.2]:57690 "EHLO peace.netnation.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1756213Ab2CBAZM (ORCPT ); Thu, 1 Mar 2012 19:25:12 -0500 Date: Thu, 1 Mar 2012 16:25:11 -0800 From: Simon Kirby To: Trond Myklebust Cc: linux-nfs@vger.kernel.org Subject: Re: [3.1-rc4] NFSv3 client hang Message-ID: <20120302002511.GA4495@hostway.ca> References: <20110909194509.GB6195@hostway.ca> <1315610322.17611.112.camel@lade.trondhjem.org> <20111020190334.GA22772@hostway.ca> <20120301225524.GB27595@hostway.ca> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii In-Reply-To: <20120301225524.GB27595@hostway.ca> Sender: linux-nfs-owner@vger.kernel.org List-ID: On Thu, Mar 01, 2012 at 02:55:24PM -0800, Simon Kirby wrote: > Here's another one, from 3.2.2. I guess it's still happening. :/ In cleaning this one up, "umount -t nfs -a -l" left the socket there still in CLOSE_WAIT. There were a few stuck "df" processes from our monitoring script: 31370 D< df -mPt nfs rpc_wait_bit_killable 31822 D< df -mPt nfs rpc_wait_bit_killable 31847 D< df -mPt nfs rpc_wait_bit_killable ...killing those made the socket actually close and change to TIME_WAIT, allowing me to remount everything, and it all works again. Simon-