| To: | netdev@xxxxxxxxxxx |
|---|---|
| Subject: | Re: [RFC] TCP congestion schedulers |
| From: | Rick Jones <rick.jones2@xxxxxx> |
| Date: | Tue, 29 Mar 2005 12:09:20 -0800 |
| In-reply-to: | <20050329120306.558cea0c.davem@davemloft.net> |
| References: | <421D30FA.1060900@ev-en.org> <20050225120814.5fa77b13@dxpl.pdx.osdl.net> <20050309210442.3e9786a6.davem@davemloft.net> <4230288F.1030202@ev-en.org> <20050310182629.1eab09ec.davem@davemloft.net> <20050311120054.4bbf675a@dxpl.pdx.osdl.net> <20050311201011.360c00da.davem@davemloft.net> <20050314151726.532af90d@dxpl.pdx.osdl.net> <m13bur5qyo.fsf@muc.de> <Pine.LNX.4.58.0503211605300.6729@dexter.psc.edu> <20050322074122.GA64595@muc.de> <Pine.LNX.4.58.0503291422100.3256@dexter.psc.edu> <20050329120306.558cea0c.davem@davemloft.net> |
| Sender: | netdev-bounce@xxxxxxxxxxx |
| User-agent: | Mozilla/5.0 (X11; U; HP-UX 9000/785; en-US; rv:1.6) Gecko/20040304 |
It could be L2 cache-coloring effects as well. Try to keep the working set size smaller than the L2 cache size of the cpu you are on. When/if using netperf, it will send from and recv to a "ring" of buffers one larger than the reported (as opposed to requested) socket buffer size divided by the send/recv buffer size. Controlled via the -W global command-line option. rick jones |
| <Prev in Thread] | Current Thread | [Next in Thread> |
|---|---|---|
| ||
| Previous by Date: | Re: [PATCH] s390: claw network device driver, Jeff Garzik |
|---|---|
| Next by Date: | [Patch] net: fix build break when CONFIG_NET_CLS_ACT is not set, Neil Horman |
| Previous by Thread: | Re: [RFC] TCP congestion schedulers, David S. Miller |
| Next by Thread: | [patch linux-2.6.11-bk10 1/1] r8169: incoming frame length check, Francois Romieu |
| Indexes: | [Date] [Thread] [Top] [All Lists] |