On 21 Aug 2002 at 14:53, Jason Corbett wrote:
> Is there an example in the code on how to do this?
It's not something you do in code - you do it in the GUI when you set up test
elements. There
are some different counters you can use, and there is a User Parameters config element
that
would allow you to create explicit values for different threads. That sounds like
what you want
and it would be easiest to use. If you have the latest JMeter code from CVS, then
fire it up,
and check out the Add-->Config Element-->User Parameters element. It works like user-
defined values, so your email address would look like:
${user}@mail.host.com
>Wow, I thought it
> was going to be alot harder to get this working, thanks. Also is it
> possible to distribute a single thread group amongst different remote
> machines. The reason I ask this is in my email stressing I need to have
> several machines running the test, but each one needs to have a
> different set of user configurations to login as. So I might want 2000
The remote testing sends the same test plan to any remote servers you want. This
means
each remote server gets a copy of the exact same test, with all its threadgroups.
This is
probably not ideal for you, unfortunately, as you'd have no way of making it different
from one
remote machine to the next.
On second thought - I wonder if you could configure your test, fire up one remote
machine,
then replace just the User Parameters element, and fire up the next remote machine?
It's kind
of diabolical to change the test while JMeter is running the test, and it might blow
up in your
face. But then again, it might let you do exactly what you want. Interesting thought.
> IMAP users, (user imap1 through imap2000), but because of the large size
> I might want to use 3 or 4 machines to distribute this across. I can
> create support for the mail stressing protocols without this, however it
> won't be able to scale up to as many users without this. I suppose
> another possiblity (though not as pretty) would be to be able to
> configure different thread groups, and run a specific thread group per
> machine, would that be easier to configure?
If my above idea doesn't work, then yes, it'd be easier to run the different machines
separately. The downside being that the data from the different machines won't get
combined.
--
Michael Stover
[EMAIL PROTECTED]
Yahoo IM: mstover_ya
ICQ: 152975688
--
To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]>
For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>
