Bram Veldhoen's OLD blog

# re: WCF Latency Test Harness and WCF Performance

Could you please provide a detailed usage and configuration guidance, perhaps also with an example? I was trying to use the code for, however failed desperately

# re: WCF Latency Test Harness and WCF Performance

I am doing similar experiment in 3.5 basically comparing basicHTTP vs WebHTTP(REST).... Have you done any similar experiment on this ?

Also I would like to know why is there so much difference b/n netpipe and basicHTTP - since in both the cases, the payload is the same (SOAP XML)....The only difference I could think of is the netpipe probably reuses the connection but not the HTTP (But when I did a similar experiment I did give the "keepAliveEnabled" for the HTTPclient class... but I still did not get any improvement... Any ideas on this?

# re: WCF Latency Test Harness and WCF Performance

Great work guys!! Validates some of my measurements without me having to do all this work!!! Thanks a million.

# re: WCF Latency Test Harness and WCF Performance

can you send WcfSvcRunner default EnterpriseLibrary.config file to me?
My Emai is zzdabing@yahoo.cn
thanks a lot

# re: WCF Latency Test Harness and WCF Performance

Can anyone tell me to configure net.null STEP BY STEP?

# re: WCF Latency Test Harness and WCF Performance

Do I understand correctly that the benchmarks were made with the client and server on the same computer? If so, have you tried the same benchmarks using two computer connected via e.g. gigabit ethernet?

# re: WCF Latency Test Harness and WCF Performance

Hi Anders,

Yes, the measurements have all been performed on one and the same computer, so no network latency involved in the measurements. However, the WCF Latency Test Harness allows you to very easily reproduce the tested scenarios, including performing the tests on two different computers (it uses SysInternals psexec to start services remotely, so you DO need the appropriate access rights on both computers).