| To: | Robert Olsson <Robert.Olsson@xxxxxxxxxxx> |
|---|---|
| Subject: | Re: pktgen |
| From: | Lennert Buytenhek <buytenh@xxxxxxxxxxxxxx> |
| Date: | Sat, 27 Nov 2004 15:39:23 +0100 |
| Cc: | hadi@xxxxxxxxxx, netdev@xxxxxxxxxxx |
| In-reply-to: | <20041127135354.GA24617@xxxxxxxxxxxxxxxxx> |
| References: | <20041124161848.GA18059@xxxxxxxxxxxxxxxxx> <16804.48120.375307.718766@xxxxxxxxxxxx> <20041124170948.GC18059@xxxxxxxxxxxxxxxxx> <16804.60621.990421.525393@xxxxxxxxxxxx> <20041125030450.GA24417@xxxxxxxxxxxxxxxxx> <16805.40983.937641.670275@xxxxxxxxxxxx> <20041127002841.GA17184@xxxxxxxxxxxxxxxxx> <20041127004325.GA17401@xxxxxxxxxxxxxxxxx> <16808.28005.74903.881087@xxxxxxxxxxxx> <20041127135354.GA24617@xxxxxxxxxxxxxxxxx> |
| Sender: | netdev-bounce@xxxxxxxxxxx |
| User-agent: | Mutt/1.4.1i |
On Sat, Nov 27, 2004 at 02:53:54PM +0100, Lennert Buytenhek wrote: > > 160 679270 > > From this point onwards your HW can saturate the pipe, this is the > boring part of the graph. Look at it this way. Assume that the cost of transmitting a single packet consists of a packet-size-dependent part (call it 'bandwidth') and a packet-size-independent part (call that one 'latency'). The higher the latter part is, the bigger packets you need to saturate the (GigE) pipe. Your 64/133 setup saturates GigE with 160B packets, my 32/66 setup needs 350B packets even though there is ample bandwidth in both cases. Hope I'm making some sense here. --L |
| <Prev in Thread] | Current Thread | [Next in Thread> |
|---|---|---|
| ||
| Previous by Date: | tcp port reuse checking TCP_LISTEN state, Ilya Pashkovsky |
|---|---|
| Next by Date: | Re: pktgen, jamal |
| Previous by Thread: | Re: pktgen, Lennert Buytenhek |
| Next by Thread: | Re: pktgen, jamal |
| Indexes: | [Date] [Thread] [Top] [All Lists] |