Hi ,
We have three jobs which we are executing sequentially in Talend Enterprise 5.4 and those are complex jobs taking lot of time for execution.
I have to ask whether every job uses the same memory block or new memory block for execution?
If same,How can we release the memory associated with that job?
Thanks in advance !!!
Hi,
What's your current job design? Are there any some cache component consuming too much memory. such as tMap, tUniqRow, tSortRow. For a large set of data, try to store the data on disk instead of memory on tMap, tUniqRow, tSortRow. Also, allocate more memory to execute the job if you have 32G memory.
Or it is Parent Job and Child job?
Please share more info about your current situation so that we can give a precise optimistic solution for you.
Best regards
Sabrina
Hi ,
job design is parent job calling 3 jobs:
job1->job2->job3
yes,there are cache components as tmap in all the jobs,and we already using store the data on disk instead of memory on tMap and allocated maximum buffer size to 25,00,000.
we have sufficient memory for our job server.,but still facing memory issues.
Can we free the memory of first job after its execution ??
Hi,
The memory should be automatically recycled in java code.
Is it working for you when you check out option "Use an independent process to run subjob" in tRunJob?(This option is not compatible with the Jobserver cache. Therefore, the execution may fail if you run a Job that contains tRunjob with this check box selected in Talend Administration Center. )
You'd better open ticket on
Talend Support Portal for your Memory Issue so that we can give a remote assistance to you.
Best regards
Sabrina
Hi ,Initially we were running the job by checking that option as run independent process but after some time of execution it was giving the error as "Child job return 137. It doesn't terminate normally.?, we are not using TAC. We are running it on unix machine.
Initially we were running the job by checking that option as run independent process but after some time of execution it was giving the error as "Child job return 137. It doesn't terminate normally.?,
Have you already opened a ticket on Talend Support Portal? Best regards Sabrina
To check whats going on I suggest you test your child job. It looks like there is something wrong with it.
The return code is the return code from the job. The message tells you that all return codes different from zero are not normally (actually it is but not for Talend).
At the moment I cannot see an error in the Talend infrastructure and therefore there is no need for a support ticket.
It would be great if you could show us some details of your jobs (screenshots etc).
To you memory questions. If you do nothing more than put the jobs into a chain, I would use an external job scheduler and not tRunJob. With a external job scheduler all jobs releases the memory at the moment they finished and not when the whole job chain has been finished.
Try SOS JobScheduler, it is free and approved.
