[SNMP4J] Memory Leak in SNMP4J (listening for traps)

[Frank Fock]

Adam,

I cannot find any leak using a profiler.

Have you forced the garbage collector to run on
a regular basis?
Have you disabled runtime-compilation in the JVM?

Best regards,
Frank

Adam Brons wrote:
> Hey Gang,
> 
> While writing an application to use SNMP4J to listen for v1, v2c, and v3 
> traps over UDP we identified a memory leak.  I'm including some log data 
> run against IBM's JVM version 1.5 as well as Sun's JVM 1.5 using the 
> SnmpRequest tool to listen for trap messages on port 162/udp. On the IBM 
> jvm SnmpRequest grew from 0.1% every 5 minutes on the Sun jvm it was 0.1% 
> every 20 minutes.  None the less, in both cases the foot print continues 
> to grow.  The command-line I was using to run SnmpRequest in both cases is 
> as follows: "java -cp SNMP4J.jar org.snmp4j.tools.console.SnmpRequest -v 1 
> -Ol 0.0.0.0/162"
> The testing was done on Linux standard PC hardware running a flavor of 
> RedHat. For my test the SnmpRequest received ZERO traps (on purpose). I've 
> yet to reduce the heap size of java such that a heap dump is performed.
> 
> FYI on the file contents I was running ps -auwwx every 60 seconds along 
> with the date command. If you have any questions let me know.
> 
>  
> 
> Adam Brons
> TSOM Software Engineer
> IBM Tivoli Software
> (404) 591-8246 office
> 
> 
> ------------------------------------------------------------------------
> 
> _______________________________________________
> SNMP4J mailing list
> SNMP4J at agentpp.org
> http://lists.agentpp.org/mailman/listinfo/snmp4j

-- 
AGENT++
http://www.agentpp.com
http://www.mibexplorer.com
http://www.mibdesigner.com