Creates so many "-APT_PMsectionLeader" sessions in

[a.thirumoorthi]

Hi ,

Whenever a parallel job is triggered,There are 100-200 sessions are opening in the Unix OS. Because of these so many sessions jobs are not running it is getting hang up.

Can any one please help on this issue , what are the possible reasons for this issues and solutions for fixing this issue.

Please help and advice if you already have encountered the same problem.

Thanks in advance

Thanks
Thiru

[throbinson]

How many nodes in your configuration file for this job run?
How many stages in the job?
What are your sources and targets?
In general and very roughly there will be a process per stage times the number of nodes defined in the configuration file.

[ray.wurlod]

You get one section leader per node per job.

Resource management is up to you - use fewer nodes or run fewer jobs simultaneously.

Contrary to popular belief, DataStage servers do not deliver an infinite supply of resources.

[a.thirumoorthi]

[quote="a.thirumoorthi"]Hi ,

Whenever a parallel job is triggered,There are 100-200 sessions are opening in the Unix OS. Because of these so many sessions jobs are not running it is getting hang up.

Can any one please help on this issue , what are the possible reasons for this issues and solutions for fixing this issue.

Please help and advice if you already have encountered the same problem.

Thanks in advance

Thanks
Thiru


Hi,

Thanks for your information.

we have configureed 4 nodes.
In my job three stages (2 Transformer, 1 joiner) are used and it is multiple instance , simulatenously two jobs are running.

Thanks
Thiru

[ray.wurlod]

You also get one process per player (operator). How many stages are in these jobs? Can you be sure no-one else is also running jobs on this server? How can you be sure?