Matthew Flanagan wrote:
> I'm trying to track down the source of a performance problem with an
> application. The application in question is from a firewall vendor and the
> component of it that is suffering is the log handling daemon. The log daemon
> is single threaded and receives logs from firewalls and also services various
> consumers which could be either simple log viewing CLI or GUI clients or log
> reporting engines running on other servers.
>
> The performance issue is seen from both the consumers and producers.
> Producers complain that the log server is not responding and therefore unable
> to send logs. The consumers don't complain (but the users do!) but are very
> slow when reading the logs, filtering on certain fields etc.
>
> I'm at a bit of a loss on where to go from here. Below is the analysis I've
> done so far. I'm looking for advice on whether this can be solved by throwing
> more resources at it or whether it is being limited by a poorly written
> application. Please let me know if you'd like to see more output from dtrace
> or other commands.
>
> Note, one thing I have seen is that if I stop the reporting engine running on
> a different server then the number of cross-calls drops significantly.
>
> The log server specs are:
>
> 4 Core T5120 with 16GB RAM
> 2TB SAN storage (mirrored ZFS pool).
> $ uname -a
> SunOS logserver 5.10 Generic_137111-01 sun4v sparc SUNW,SPARC-Enterprise-T5120
^^^^^^^^^
First question (+generic suggestion): Did you ever apply the Recommended
Patch Cluster ? AFAIK the latest kernel patch is 137137-09 (see
http://sunsolve.sun.com/search/document.do?assetkey=1-21-137137-09)
which should fix some issues related to high number of crosscalls etc.
----
Bye,
Roland
--
__ . . __
(o.\ \/ /.o) [email protected]
\__\/\/__/ MPEG specialist, C&&JAVA&&Sun&&Unix programmer
/O /==\ O\ TEL +49 641 3992797
(;O/ \/ \O;)
_______________________________________________
perf-discuss mailing list
[email protected]
