On 17/10/2008, Peter Lin <[EMAIL PROTECTED]> wrote: > the key here is to determine was the peak request per second is. 24000 > users that are inactive might only generate 100 request/second. 1000 > very active users could easily exceed the load generated by 24000 > inactive users. >
True, but the total number of sessions stored in a server may affect its performance, so it may be necessary to generate many different session ids. But I agree that the starting point is to look at the requests per second. If the server cannot handle the target load with fewer sessions, then it is not going to cope with the same load generated by more sessions. > > peter > > > On Fri, Oct 17, 2008 at 8:18 AM, rustix <[EMAIL PROTECTED]> wrote: > > > > i've created a wap application using java and wml. this system has to be > able > > to support 24000 concurrent sessions at a given time. how could i carry > this > > out? i'm hoping to use JMeter > > -- > > View this message in context: > http://www.nabble.com/-Help--Checking-number-of-concurrent-session-with-JMeter--tp20032128p20032128.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] > > --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
