Page 1 of 1

Restarting Datastage Services

Posted: Thu Jun 07, 2007 8:45 am
by rexsmith
I had my job locked especially RTCONFIGnnn file. Our Datastage Server went to routine maintainence on the same evening. They normally restart the DS engine but to my surprise i had my job still locked after the reboot .I had to ask admin to unlock the files. Some body can tell me what could be the reason?

Posted: Thu Jun 07, 2007 11:33 am
by battaliou
Stopping and starting your datastage services will not necesserally unlock your jobs. This is because the datastage client server connect is persistent. What you need to do at UNIX is:
netstat -na|grep 31538

This will tell you which processes are connected to the engine and what state they are in. If they are in a CLOSE_WAIT state, then wait a while and they should die after a few minutes. Of course, if you're sure everyone has logged of, login into the uv account of datastage as dsadm and go:
UNLOCK ALL

I've never had a problem using this command, but use it wisely.

Posted: Thu Jun 07, 2007 1:25 pm
by kcbland
There's no way your job was locked after a reboot of the server. I suspect the server was not rebooted (use uptime command to verify) and that the services were never recycled (ps -ef|grep dsprpcd and look at the start time of the daemon).

Next time you have a lock, use the UNLOCK USER nnnn ALL command from the DS Engine project (UV) to release locks for your job number. Search the forum for more history about this.

Posted: Fri Jun 08, 2007 1:48 am
by ray.wurlod
A lock on RT_CONFIGnnn usually indicates that someone has a Monitor open on job number nnn. This will properly prevent the job from being recompiled, which would overwrite the RT_CONFIGnnn table from which the Monitor is obtaining some of its information.