[Orca-users] Re: Memory leak in orcallator.se

[Blair Zajac]

Hi Guilherme,

Which orcallator.se are you using?

Does iostat.se show the same increase in memory usage if you leave it
running for a long time?

If so, I would send the problem to se-feedback at setoolkit.com as that
where SE specific issues would best be discussed.

If it doesn't happen on iostat, then if you could determine which
portion of orcallator.se has the memory problem, that would help
a lot.

Best,
Blair

Guilherme Carvalho Chehab wrote:
> 
> I have been experiencing a very strange thing on one of my servers
> when running
> orcallator.se. When it is started it shows the following error
> messages:
> 
> # /etc/init.d/orcallator start
> Writing data into /usr/local/var/orca/orcallator/orca/
> Using www access log file /usr/local/apache/logs/access_log
> Starting logging
> couldn't map disk instances
> couldn't obtain disk partition usage
> 
> Regardless, it continues to run for some hours, while it is running
> the system
> memory usage grows very quickly until all available RAM (and swap
> area, of
> course) are depleted. At this point se fails with a "not enough
> memory" message.
> 
> I saw someone on the list had the same messages on failure obtaining
> disk map
> and partitions instances. But he didn´t mentioned anything regarding
> memory
> problems.
> 
> Using the iostat.se distributed with SE Toolkit 3.2 I noted that the
> fault
> behavior was the same. The server is Ultra 10 (IDE disks) a Solaris 8
> upgraded
> from 2.6 (a lost library, maybe?). And everything else run just fine
> on the
> machine.
> 
> Have anyone had the same problem?
> 
> regards
> Guilherme