[ic] IC process memory and CPU usage growing

Kevin Walsh kevin at cursor.biz
Thu Oct 9 14:17:57 EDT 2003


Kevin Old [kold at kold.homelinux.com] wrote:
> I'm running IC 4.9.7, MySQL 3.23, 512MB RAM, 1Ghz processor.  IC has
> been running just fine, but today I noticed that the IC process is
> roughly using 36%-99.9% of the CPU and 28M-58M of RAM.
> 
> I'm only running one catalog and this is not a very high traffic site,
> although I have it configured as High in interchange.cfg.
> 
> I've cleaned out the tmp directory under the catalog directory and
> expired all the sessions under that directory in the catalog directory.
> 
> Any ideas what could be causing this?
> 
> No errors in either error.log.
> 
> Even with all this the site seems to respond pretty fast.  I'm just
> concerned that it'll disable my server.
> 
What has changed between the time the system was running ok and the
time you noticed the high CPU usage?  For instance, have you
upgraded Interchange, Perl or any Perl modules?  Have you modified any
of your pages in a way that you can determine what as changed there?

What operating system, and O/S version, are you using?  Also, what
Perl version is in use and how was it built?

Does the memory usage grow as you watch the high CPU process or does
it remain static?  You said that the memory usage was 26-58MB.  What
is normal for your setup?

Have you tried running "strace -p 1234", where 1234 is the process ID,
to determine whether the process is spinning or simply doing some
heavy-duty work?  If so, can you post a brief section from the end of
the trace if it indicates that the process is stuck in a tight loop.

Does the process settle down after a while, or does it remain using
a lot of CPU time until it is killed?

Lastly, is the process killable, or does it require a "kill -9" to
get rid of it?  If you attempt to stop Interchange and then do a "ps",
can you see that the high-CPU process has died, or does it linger
until it is specifically killed?

I have received three reports of this sort of thing, two from the same
person but on different machines.  Yours is the forth report I've seen,
if it turns out to be the same problem.

I have, so far, been unable to determine a cause, or find a coincidence
in each of the setups that may be to blame for this.  Answers to the
above questions may offer a clue.

If anyone else is suffering from a similar problem then please post as
much detail as you can gather - short of a 100MB strace log. :-)

-- 
   _/   _/  _/_/_/_/  _/    _/  _/_/_/  _/    _/
  _/_/_/   _/_/      _/    _/    _/    _/_/  _/   K e v i n   W a l s h
 _/ _/    _/          _/ _/     _/    _/  _/_/    kevin at cursor.biz
_/   _/  _/_/_/_/      _/    _/_/_/  _/    _/



More information about the interchange-users mailing list