From mboxrd@z Thu Jan 1 00:00:00 1970 From: ebiederm@xmission.com (Eric W. Biederman) Subject: Re: RFC Hanging clean-up of a namespace Date: Sun, 22 Jan 2012 22:25:52 -0800 Message-ID: References: <20120119192541.GM2262734@jupiter.n2.diac24.net> <201201201251.25032.hans.schillstrom@ericsson.com> <201201230707.33761.hans.schillstrom@ericsson.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: Hagen Paul Pfeifer , David Miller , "equinox\@diac24.net" , "netdev\@vger.kernel.org" To: Hans Schillstrom Return-path: Received: from out02.mta.xmission.com ([166.70.13.232]:57748 "EHLO out02.mta.xmission.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751677Ab2AWGXX (ORCPT ); Mon, 23 Jan 2012 01:23:23 -0500 In-Reply-To: <201201230707.33761.hans.schillstrom@ericsson.com> (Hans Schillstrom's message of "Mon, 23 Jan 2012 07:07:32 +0100") Sender: netdev-owner@vger.kernel.org List-ID: Hans Schillstrom writes: > On Friday 20 January 2012 21:55:27 Eric W. Biederman wrote: >> My current hypothesis is that the namespace actually didn't get freed >> until the tcp socket finished closing. You can check by looking at when >> __put_net and then cleanup_net are called. > > __put_net() is called just after tcp_write_timer() fires and then > cleanup_net() Hypothesis confirmed. Your speed problem is that it is taking 2 minutes in the pathological case for your tcp socket to close. Do you have any clue why it is taking your sockets so long to close? Is the other side simply not responding? Eric