2008/4/21 afsilva <[EMAIL PROTECTED]>: > > > > > sebb-2-2 wrote: > > > > On 5/31/05, Yuval <[EMAIL PROTECTED]> wrote: > >> Hi, > >> I running a load test with the TCP sampler, and I'm experiencing a > >> strange > >> behavior: > >> It seems like the minimum response time for a TCP sampler is always the > >> value of the TIMEOUT (which is used in *setSoTimeout* in the code). > >> Doesn't this value suppose to be a maximum value for a *read* to be > >> performed? > >> Thanks, > >> Y. > >> > > Have you set the eol character? > > It needs this to know when the response is complete ... > > > > > > Hi, I am experiencing the same beaviour but I'm sending the eol char, I mean > byte. > > on the jmeter log I can see the output: > > ... > 2008/04/21 17:01:01 INFO - jmeter.protocol.tcp.sampler.TCPClientImpl: Using > eolByte=0 > ... > > my tcp message ends with that byte, but the response is still getting only > after the timeout is reached. > > Am I missing something? >
Yes - the code ignores the eolByte if it is zero ;-) It was unfortunate that a valid byte value was chosen to mean ignore EOL checking. It was assumed that a zero byte was very unlikely to be used as an EOL byte. > -- > View this message in context: > http://www.nabble.com/Socket-Timeout-tp105444p16808257.html > Sent from the JMeter - User mailing list archive at Nabble.com. > > > --------------------------------------------------------------------- > To unsubscribe, e-mail: [EMAIL PROTECTED] > For additional commands, e-mail: [EMAIL PROTECTED] > > --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
