On Sunday, Apr 4, 2004, at 14:53 US/Mountain, Jason R. Mastaler wrote:
Jeff Ross <[EMAIL PROTECTED]> writes:
I know that is your belief, Jason
And that's only because:
1) tmda-ofmpid was working fine for you until you underwent a major OS and hardware upgrade of the system.
Not sure about the "major" part. I went from OpenBSD 3.3 to 3.4 (a.out -> ELF, though) and then to 3.5-current. Hardware went from a PI/120 or thereabouts with 48 MBs of ram to a Pentium Pro 200 MHz with 128 MBs of ram. My "state-of-the-art" e-mail server ;-)
2) No one else has reported this problem.
3) I don't see any problems in the code that would cause this.
I'm always open to being proved otherwise though. This may just be a really hard bug to track down.
Nothing else in my mailserver shows any other anomalies.
Are there any other long-running Python processes on the machine such as Mailman? Otherwise, this is an apples and oranges comparison.
Good point. My only experience with Python has been with TMDA, and I've always thought it was a bit piggish in memory use, but well worth it! Watching top, whenever I see it kick in for a user I see a couple of python processes, each using 11 to 13 MBs or ram. Is that normal?
Other than Mailman, are there any other python programs you can think of that might give me a little torture test? Browsing the Python web pages led me to try CAGE, a cellular automaton program. It runs at near 100% CPU, using between 2 and 4 MB of ram, depending on the generation algorithm.
PID USERNAME PRI NICE SIZE RES STATE WAIT TIME CPU COMMAND 28169 jross 64 0 2692K 3560K run - 20:50 98.58% python
It's a pretty quiet night on the e-mail front, so I'll let this run for a while and see what happens. As I near 15 minutes, though, it hasn't used any more ram than it did at startup.
I also saw a program called tracy that is supposed to do a trace of live running Python programs, but I have yet to figure out how to use it. Is there any other way I can take a look see at how the tmda-ofmipd process is running here?
I don't do virtual users, so I can't help the OP with that aspect of his tmda-ofmipd problem.
Until I figure out what's going on whenever I see a problem being reported about tmda-ofmipd, I'm going to bring the issue up.
That's fine as long as it's relevant to the problem being reported. Most "problems" are really just configuration issues.
Thanks,
Jeff _________________________________________________ tmda-workers mailing list ([EMAIL PROTECTED]) http://tmda.net/lists/listinfo/tmda-workers
