On 14.04.23 04:55, pratik.kulka...@shell.com.INVALID wrote:
Hi All,

Thanks for your quick suggestions.

As Olaf suggested, I tried to set the same values to Xms and Xmx; the 
application immediately crashes after Tomcat restart and I am not able to 
access it. I tried it with different values, but the result is the same. 
Regarding the memory leak, let me check on how to monitor an Oracle Apex 
application running on Tomcat, since this is something I have never done before.

Here are a few ideas for this paragraph:

As your application can't run with -Xms set to the limit, you have proven that your computer does not have enough memory to allocate for your requirements (maybe due to other applications taking it, maybe it's just due to low RAM to begin with. I'm not sure why this would surface after upgrading Tomcat, but maybe your settings are different in other aspects as well, or other applications now take more memory.

(I'd recommend Xms==Xmx for all Java Server applications, to get the OOME quicker, rather than Sunday night at 3am, at random)

If you just connect jconsole, you should be able to see a characteristic sawtooth curve of memory, going up, hinting at - not necessarily proving - memory leaks (it might also just indicate legitimate memory usage - it's a quick check: If the sawtooth is horizontal, you have a strong indicator that there's no memory leak)

Typically, I don't really concentrate on the actual stacktrace of the OOME, as this is just what happened to require the last bit of memory, after everything else has been consumed. This *might* be the guilty code, but it might also just be a false positive. To me, OOME stacktraces have more often misled me than they pointed to the problem. A heapdump, with an indicator of what actually consumes your memory (and then judging if it's legit or not) helps a lot more IMHO.

Olaf


---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscr...@tomcat.apache.org
For additional commands, e-mail: users-h...@tomcat.apache.org

Reply via email to