Re: Using udpgen.click without polling device patch???
shule ney <[email protected]>
| Newsgroups | gmane.network.routing.click |
|---|---|
| Message-ID | <CAF8yJrOhS6bNx9JLhp88GPUiho8_BOJsHOdntvEtL=S2ZzhjhA@mail.gmail.com> |
Thanks for you reply Eddie, the problem is at first I used udpgen which can sent 64B packet at about 1.3Mpps, but after I reinstall the linux on the same machine, it can only achieve around 750Kpps which is not fast enough(CPU=100%). I wonder where is the problem, the polling driver patch?? when I start udpgen I can see this "*ToDevice eth1 is full, packet delayed* ". 在 2011年9月28日 下午9:55,Eddie Kohler <[email protected]>写道: > Hi Shule, > > 1. Use FromDevice instead of PollDevice. It will be slower, but might still > be > fast enough for you. > 2. Yes, if you just leave the line in, I would expect 0 rate: the > PollDevice > does nothing. > > Eddie > > > On 9/27/11 9:58 PM, shule ney wrote: > > *HI all:* > > The udpgen.click is as below: > > > > // UDPGen($device, $rate, $limit, $seth, $sip, $sport, $deth, $dip, > $dport); > > // > > // $device name of device to generate traffic on > > // $rate rate to generate traffic (packets/s) > > // $limit total number of packets to send > > // $size bytes per packet > > // $seth source eth addr > > // $sip source ip addr > > // $sport source port > > // $deth destination eth addr > > // $dip destination ip addr > > // $dport destination port > > > > elementclass UDPGen { > > $device, $rate, $limit, $size, > > $seth, $sip, $sport, $deth, $dip, $dport | > > > > source :: FastUDPSource($rate, $limit, $size, $seth, $sip, $sport, > > $deth, $dip, $dport); > > * **pd :: PollDevice($device) -> ToHost;* > > source -> td :: ToDevice($device); > > } > > > > // create a UDPGen > > > > u :: UDPGen(eth1, 82000, 820000, 1500, > > 00:e0:29:05:e2:d4, 2.0.0.2, 1234, > > 00:c0:95:e2:09:14, 1.0.0.2, 1234 > > > > 1.So what does pd used for??? If I don't have polling device patch, > > will it send much slower?? > > > > 2.When I don't commented out this line, it also works, and the rate of > > u is 0, is it common??? > > _______________________________________________ > > click mailing list > > [email protected] > > https://amsterdam.lcs.mit.edu/mailman/listinfo/click > _______________________________________________ click mailing list [email protected] https://amsterdam.lcs.mit.edu/mailman/listinfo/click