> You will need to have the service called something different than the
server
> name so traffic can be routed correctly - probably
jud.vkhdli08.hda.hydro.com.
> This should be set both in the service section for jud, and the agents
block
> within the jsm 'sessions' configuration (the former telling the server how
it
> should contact users-agent, the latter telling clients that the jud exists
and
> where). There may be a requirement that jud.vkhdli08.hda.hydro.com be
> resolvable, if the users-agent uses a socket for the connection - in that
case
> I'd recommend tossing that in the /etc/hosts file. Clients will speak to
> jud.vkhdli08.hda.hydro.com through the server, so they won't need to know
the
> IP themselves.

If there is a requirement that the service be named different than the
server, it's all ok by me.
However, where is this documented? I did not see anything in the howto or
install docs for the JUD.

I'll test it tomorrow.

Best regards
Ørjan Sandland



I am working with a company that will be using Jabber. I have been
assigned the task of determining performance numbers for issues of
scalability. In order to do this I have been working on a suite of
programs to test various aspects of the Jabber server's (or platform's!)
performance and limitations. Specifically, here are the numbers we want:

o limitation: maximum connections
o performance, passive, for n users:
        * time to connect
        * time to establish stream
        * time to login
o performance, active, for n users:
        * time to connect
        * time to establish stream
        * time to login
        * time to send x messages between y users

The difference between passive and active is that with passive the n
clients connect and then wait. With the active test each client will
connect and then begin sending some arbitrary number of messages to
other clients while the new clients in turn connect and do the same. I
have pretty much completed the limitation test and passive performance
test. The next stage is the active performance testing. BTW, we are
aware of simulate3. However, I didn't find it very effective for our
needs. Also, the source's documentation was rather terse, to say the
least.

Is anyone interested in using this suite as well as helping to work out
any problems? We will be releasing the test suite as GPL'd software
available on SourceForge. We would be VERY happy if members of jdev and
jadmin would checkout the code and make any suggestions they felt would
make the software more effective. 

I have just created the SourceForge account and project, so it may take
up to two days for the project to become active.

Regards, Dustin

-- 
Dustin Puryear <[EMAIL PROTECTED]>
http://members.telocity.com/~dpuryear
In the beginning the Universe was created. 
This has been widely regarded as a bad move. - Douglas Adams

_______________________________________________
jdev mailing list
[EMAIL PROTECTED]
http://mailman.jabber.org/listinfo/jdev


Reply via email to