From mboxrd@z Thu Jan 1 00:00:00 1970 From: Stephen Hemminger Subject: Re: [net-next PATCH 3/3] qlge: Increase default TX/RX ring size to 1024. Date: Thu, 11 Jun 2009 08:34:19 -0700 Message-ID: <20090611083419.4c173558@nehalam> References: <1244684975-10211-1-git-send-email-ron.mercer@qlogic.com> <1244684975-10211-3-git-send-email-ron.mercer@qlogic.com> <20090611.022713.35602614.davem@davemloft.net> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Cc: ron.mercer@qlogic.com, netdev@vger.kernel.org To: David Miller Return-path: Received: from mail.vyatta.com ([76.74.103.46]:37484 "EHLO mail.vyatta.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752490AbZFKPeW (ORCPT ); Thu, 11 Jun 2009 11:34:22 -0400 In-Reply-To: <20090611.022713.35602614.davem@davemloft.net> Sender: netdev-owner@vger.kernel.org List-ID: On Thu, 11 Jun 2009 02:27:13 -0700 (PDT) David Miller wrote: > From: Ron Mercer > Date: Wed, 10 Jun 2009 18:49:35 -0700 > > > > > Signed-off-by: Ron Mercer > > This is huge. Even other aggressive NICs such as BNX2X only use 256 > ring entries per TX queue. > > There is a point where increasing definitely hurts, because you're > increasing the resident set size of the cpu, as more and more SKBs are > effectively "in flight" at a given time and only due to the amount > you're allowing to get queued up into the chip. > > And with multiqueue, per-queue TX queue sizes should matter less at > least to some extent. > > Are you sure that jacking the value up this high has no negative side > effects for various workloads? I am investigating reducing the sky2 default tx ring size down to 128 after user complaints about the latency. At 10G 1024 ring entries is 7ms for jumbo frames.