jimmy35 wrote:
Bernd Fondermann wrote:
jimmy35 wrote:
Bernd Fondermann-2 wrote:
The users should get created on the server specified by <testserver
host="..." >
Yes, this has to be the IP/name of the tested server.
I changed this to match the external server, but it failed,
What does that mean, it failed? Does it throw an exception? Is the
postage machine able to connect to the remote manager port on the other
machine? Please give more information, otherwise it will be hard to help
you out.
I've since
posted the scenario in a txt file.
That looks fine. Maybe you can copy Postage over to the other machine
and check again.
The test is running that long despite a 10 minute setting in the
config.xml. It may be that we're exceeding the number of emails we can
process and that is causing the test duration to be that long. I do
notice that the James java process starts to churn through memory during
a
test of this nature. The output log does list all the emails but the
bottom portion just echoes my param settings (i.e. 3000 emails/min,
30000
emails sent, etc...)
Mmhhh. I have no idea what the problem is here. Do you see any activity
in the James logs after 10 minutes?
I will check the latest logs regarding the activity. It seems that we are
sending now at the correct speeds, and someone had been using fetchmail
which was distorting our test run.
With today's multicore and many GB-RAM machines it is very likely that
the numbers from the JamesByTheNumber wiki page do no longer apply.
If you have new findings, please let us know! Thank you very much!
I will post whatever settings we find relevant, along with the machine
specs, once we get this running as we need it to. Right now we're seeing
bounce1 send all the mails to bounce2, but bounce2 is eating up memory
until it makes the machine unresponsive, so I am assuming there is
something wrong with either our scenario or our config.xml for James.
Cory
BTW, could you please fix you email client to not mark your answers as
quoted? It's hard for me to track what you added and what is from
previous mails.
Bernd
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
