On Thu, 16 Dec 2004 10:35:38 -0000, Ashley Hayes
<[EMAIL PROTECTED]> wrote:
> My mistake, I miss interpreted my results ( my iteration can take quite a
> long time and was longer that the test duration ).
> So I can confirm that a start delay of 1 works for me, even with the start
> and end times as their default.
>
> For others out there the answer to my original question is that version
> 2.0.2 expects a Startup delay of > 0 to consider the Duration parameter.
Not quite:
JMeter 2.0.2 calculates the end-time from the start-time + the
duration (if > 0).
So if the startup-time is now or in the future, the duration works as expected.
However, if the startup time is in the past, a startup delay of > 0 is
needed to make sure that the startup time is reset before calculating
the end time.
This has been fixed, so that the next version of JMeter (2.0.3?) will
assume that a past start-time means now.
>
> Thanks sebb/peter,
> A
>
> P.s.
> My new config is:
> Start Time is the default "1970/01/01 01:00:00"
> End Time is the default "1970/01/01 01:00:00"
These are only the defaults when loading an existing test plan which
does not have any times set. This should probably be changed to assume
current time/date.
If a thread group is created in a new plan, the defaults are the
current time/date.
> Duration (seconds) is ${TestDurationInSeconds}
> Startup delay (seconds) is 1
>
>
[...]
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
