I believe the constant timer uses milliseconds for the delay. the way it works is once a response is done, it will wait x milliseconds before starting another request. does that answer your question? peter On Mon, 5 Jul 2004 03:04:17 -0700 (PDT), shambu pujar wrote: > > hi > > whether Constant timer delay expressed ,in milli sec,is between response time of last request and the request to be made or is it between the requests??? > > we ran 12500 users in 21 m/cs with constant delay timer of 30 mins.if u trace a thread from single m/c raw data output file, then i found that delay of 30 min is between response time of the last request made by that thread and the request time of the present request. This was true in many cases!.but after 1 1\2 hour through the test , the delay was over 30 min by 49 sec or so with many threads in all machines.It was not 30 min constant. At this moment many request are sent in a sec, creating to raise in cpu utilization of server and also response time increased.The test lost its stability later onwards. > > i feel timer is not able to time the request exactly ???? for first 1 1'2 hours it went smoothly??this has been the case for all the 3 tests we had run on different days??? could any one help me what went wrong in test ????my test script is simple! i m sending three requests. timer inserted at first request! so after making three requests thread has to wait for 30 min to make the first request again! test is run for 2 1'2 hours. > > regards > > > --------------------------------- > Do you Yahoo!? > Yahoo! Mail is new and improved - Check it out! --------------------------------------------------------------------- To unsubscribe, e-mail: jmeter-user-unsubscribe@jakarta.apache.org For additional commands, e-mail: jmeter-user-help@jakarta.apache.org