that wouldn't ensure clean restart after JVM crashes though (or hardware failure, which causes an abnormal JVM termination).
For that you need some sort of mechanism for the application to constantly keep track of where it is and a way to continue halfway through a compound unit of work.
> sounds like he wants something that monitors his
> application and restarts it automatically when it
> crashes.
I agree.
@Op. Google on process watchdog.
>
> Maybe something like Nagios combined with a
> shellscript to parse the output and a cronjob to run
> that shellscript every few minutes?
Nagios ca be good, but at watchdog is probably better than the cron job.
Sorry i used the word failover incorrectly. Thanks for correcting n understanding the prob.
I dont know whether i should use any monitoring tool or not, can u suggest me any other way. Can i use JMX here? If i use JMX then iam afraid i have to make changes in the existing code.
Pls advise.
> Sorry i used the word failover incorrectly. Thanks
> for correcting n understanding the prob.
> I dont know whether i should use any monitoring tool
> or not, can u suggest me any other way. Can i use JMX
> here? If i use JMX then iam afraid i have to make
> changes in the existing code.
> Pls advise.
Use a process watchdog.