Job Aborted: Resource temporarily unavailable

Post questions here relative to DataStage Enterprise/PX Edition for such areas as Parallel job design, Parallel datasets, BuildOps, Wrappers, etc.

Moderators: chulett, rschirm, roy

Post Reply
apraman
Participant
Posts: 47
Joined: Mon Sep 12, 2005 5:26 am

Job Aborted: Resource temporarily unavailable

Post by apraman »

Hi,
I have parallel job. While run time the job get aborted. Some of the fatal error messages are -

Code: Select all


main_program: APT_PMConnectionRecord::rsh: fork failed, Resource temporarily unavailable

main_program: APT_PM_LocalShell: fork failed, errno = 11

main_program: The section leader on gifbdc died


-----------------------------------------------------

node_node1: Fatal Error: Unable to start ORCHESTRATE process on node node1 (gifbdc): APT_PMPlayer::APT_PMPlayer: fork() failed, Resource temporarily unavailable
How to overcome this situation?
Arun
ArndW
Participant
Posts: 16318
Joined: Tue Nov 16, 2004 9:08 am
Location: Germany
Contact:

Post by ArndW »

Arun,

the operating system is trying to execute a fork() and can't. The fork() call is used to spawn new processes from a parent {it actually duplicates a processes' complete memory space and starts a duplicate process}.

If this error happens every time you run this job then most likely you have hit a [soft] limitation on the number of concurrent processes that one userid may have (look up "maxuprc" for your version of UNIX). You could have your system administrator raise that limit. All see your manpages for nproc or max_nprocs .

It might be that your system is being overloaded with requests; UNIX is usually very good about fork()ing; so it would have to be very busy for this to be the cause.
kumar_s
Charter Member
Charter Member
Posts: 5245
Joined: Thu Jun 16, 2005 11:00 pm

Post by kumar_s »

HI,

As Arnd suggested you need to check the number of Concurrent process, and tune the unix settings.
Also check the number of jobs executed parallely by the time of the error. Its obviuos that the number of jobs increased the number of process. If this is the case, you can restrice the number of jobs execution easily rather than controling the umber of processes.

-Kumar
Post Reply