From mboxrd@z Thu Jan 1 00:00:00 1970 From: Stephen Hemminger Subject: Re: [IPV4 0/9] TRIE performance patches Date: Thu, 24 Jan 2008 08:18:21 -0800 Message-ID: <20080124081821.018476cd@deepthought> References: <20080122233733.404145234@linux-foundation.org> <18327.18935.137799.285515@robur.slu.se> <20080123154933.233c4909@deepthought> <18328.23597.205001.332244@robur.slu.se> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Cc: Robert Olsson , David Miller , netdev@vger.kernel.org To: Robert Olsson Return-path: Received: from smtp2.linux-foundation.org ([207.189.120.14]:54796 "EHLO smtp2.linux-foundation.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752642AbYAXQV4 (ORCPT ); Thu, 24 Jan 2008 11:21:56 -0500 In-Reply-To: <18328.23597.205001.332244@robur.slu.se> Sender: netdev-owner@vger.kernel.org List-ID: On Thu, 24 Jan 2008 10:36:45 +0100 Robert Olsson wrote: > > Stephen Hemminger writes: > > > Dumping by prefix is possible, but unless 32x slower. Dumping in > > address order is just as logical. Like I said, I'm investigating what > > quagga handles. > > How about taking a snapshot to in address order (as you did) to some > allocated memory, returning from that memory in prefix order? This would > solve the -EBUSY too and give a consistent view of the routing table at > the time for the dump/snapshot. > > Cheers > --ro Snapshotting is going to work, because of scale and because the kernel can't tell when application is going to come back. -- Stephen Hemminger