Adding below to the wso2server script can solve the issue.
*time_stamp=$(date +"%Y.%m.%d.%H.%M.%S")*
while [ "$status" = "$START_EXIT_STATUS" ]
do
$JAVACMD \
-Xbootclasspath/a:"$CARBON_XBOOTCLASSPATH" \
$JVM_MEM_OPTS \
-XX:+HeapDumpOnOutOfMemoryError \
-XX:HeapDumpPath="$CARBON_HOME/repository/logs/heap-dump-*$time_stamp*.hprof"
\
$JAVA_OPTS \
-Dcom.sun.management.jmxremote \
Thanks,
Chamara Silva
On Thu, Mar 8, 2018 at 9:51 PM, Lahiru Sandaruwan <[email protected]> wrote:
> +1 Chamara. This has been an issue for several users as I have also
> experienced. However, we need to advice the users to delete the older heap
> dumps as the server disk space may run out.
>
> Thanks.
>
> On Thu, Mar 8, 2018 at 7:35 AM, Chamara Silva <[email protected]> wrote:
>
>> According to the current implementation, If the server goes to OOM, heap
>> dump will not create if log directory already has the old heap dump file.
>> IMHO we need to manage this by creating heap dump file with the different
>> timestamp. Otherwise, we have to manually delete them. The impact is, some
>> customer's faced same situation and we cannot help them due to the
>> unavailability of latest heap dump.
>>
>> Thanks,
>> Chamara Silva
>> --
>> Suminda Chamara Silva
>> Technical Lead
>> WSO2 Inc.
>> Mobile: +94718302858 <071%20830%202858>
>> blog: http://chamaras.blogspot.com
>>
>>
>> _______________________________________________
>> Dev mailing list
>> [email protected]
>> http://wso2.org/cgi-bin/mailman/listinfo/dev
>>
>>
>
>
> --
> --
>
> Lahiru Sandaruwan
> WSO2 Inc., http://wso2.com
>
> lean.enterprise.middleware
>
> m: +1 901 530 2379 <+1%20901-530-2379>
> e: [email protected] b: https://medium.com/@lahirugmg
> in: http://lk.linkedin.com/pub/lahiru-sandaruwan/16/153/146
>
>
--
Suminda Chamara Silva
Technical Lead
WSO2 Inc.
Mobile: +94718302858
blog: http://chamaras.blogspot.com
_______________________________________________
Dev mailing list
[email protected]
http://wso2.org/cgi-bin/mailman/listinfo/dev
