DSXchange

Hi, Our daily batch job aborted today with a message:
Resource temporarily unavailable.

The daily job has been running for a year and this message is received first time today.
Found a related link on this topic
http://www-01.ibm.com/support/docview.w ... wg21645480

Indicates the cause to be "insufficient user limits for number of processes or file handles, on UNIX/Linux systems, or problems with the job monitor processing for a job.
Using ulimit -u command found file handles and processes are limited to 4096.

1) Does that mean the number of processes are going beyond 4096 hence causing the abort?
2) Do we need to increase the processes as said in the site?

Short answers? Yes. Yes.

Longer answer...

The tech note does state: "set both the user process limit (ulimit -u) and file handle limit (ulimit -n) to 10000 or more."

So insure that both "nofile" and "nproc" are set appropriately.

I was hoping that was covered by the "as said in the site" comment.

So the steps would be ,add the following entries in dsenv file
ulimit -u 10000
ulimit -n 10000

Also ,edit the following entry under /etc/security/limits.d/20-nproc.conf file as below:
* soft nproc 10000

what is the path for nofile ?

You dont need to edit the dsenv

Depends which operating system you have.

For RHEL, you can edit /etc/security/limits.conf
dsadm soft nproc 10000
dsadm hard nproc 10000


After doing this, logout, relogin as dsadm (or whatever user you use for datastage)
Execute the shell command:
ulimit -a

check if new values are propagated

skathaitrooney,

Thanks a lot .

So... resolved?