[Geoserver-users] Apache Tomcat crashing...

A short update: I think I was able to reproduce the crash. It happened after trying to reproject the data from the native srs to epgs:4326. The requested wms layer is cascaded in MapServer with aerial photographs (to obtain a google like appearance). It does not seem to crash with requests in the native srs.

Stefan

-----Ursprüngliche Nachricht-----
Von: Andrea Aime [mailto:aaime@anonymised.com]
Gesendet am: Montag, 18. August 2008 12:01
An: Ziegler Stefan
Cc: geoserver-users; mika
Betreff: Re: [Geoserver-users] Apache Tomcat crashing…

Ziegler Stefan ha scritto:

Swap space looks like this:
total used free shared buffers cached
Mem: 8306064 6183208 2122856 0 60492 5620972
-/+ buffers/cache: 501744 7804320
Swap: 4096532 656 4095876

Thanks for the link. The “Troubleshooting and Diagnostic
Guide” seems to
be full of information but it’s getting a little bit late for this
hardcore java action (and a new episode of
“Californication” will be
aired in half an hour ;)).

Looking at the page Jukka linked I see this:

“In Mustang this condition will trigger the VM to invoke the
fatal error
handling mechanism. This means you should get a fatal error
log as you
would get with a normal (abnormal?) crash. The fatal error
log is named
hs_err_.log and contains useful information about the thread,
process, and system at the time of the crash. In the case of
native heap
exhaustion then the heap memory and memory map information in the log
can be useful. The exact format is version and platform
specific and you
will get more information in the J2SE Troubleshooting Guide.”

Mumble, can we see the contents of that hs_err_.log file?
Cheers
Andrea

Ziegler Stefan ha scritto:

A short update: I think I was able to reproduce the crash. It happened after trying to reproject the data from the native srs to epgs:4326. The requested wms layer is cascaded in MapServer with aerial photographs (to obtain a google like appearance). It does not seem to crash with requests in the native srs.

There are probably two things at work here. The first is a solid bug
in the Sun JVM, if the shape is problematic, it should report an
error, not crash like this. Trying to upgrade the JVM _might_ solve
the issue.

Besides this, maybe the shapefile renderer is being too aggressive in
trying to simplify the geometry before rendering it. I've fixed
a related issue recently (mid July), you might want to grab a
nightly here: http://gridlock.openplans.org/geoserver/1.6.x/
and see if that fixes your issue.

That release is quite close to being turned into 1.6.5.
Cheers
Andrea