Encountered the problem that the job status is stuck at running for more than an hr... even though I already set the constraint to extract and process only 1 record.
There could be many possibilities but is there any of those common causes that anyone know of?
How can I trace the job record by record? Currently the Director only shows stages of the job.
Thanks in advance! [:)]
Job Hangs When Running
Moderators: chulett, rschirm, roy
Hi,
few ideas you might wana check out, in case your problem happens whenever you run your job:
1. is there any long checkpoint or rollback on hte DB server @ the time your job was stuck?
2. are you using a routine you wrote that might contain an infinite loop?
3. are you locking something and trying to use it again before releasing it? (like DS LOCK mechanisim or using seq file not closing it and trying to open/read it again in batches etc...)
Roy R.
few ideas you might wana check out, in case your problem happens whenever you run your job:
1. is there any long checkpoint or rollback on hte DB server @ the time your job was stuck?
2. are you using a routine you wrote that might contain an infinite loop?
3. are you locking something and trying to use it again before releasing it? (like DS LOCK mechanisim or using seq file not closing it and trying to open/read it again in batches etc...)
Roy R.
-
- Premium Member
- Posts: 483
- Joined: Thu Jun 12, 2003 4:47 pm
- Location: St. Louis, Missouri USA
nelc,
Just wanted to pass along my findings with a similar situation... I had this exact situation and I found that I was doing a lookup on an external database (Oracle) and didn't give it enough fields to use the correct index, causing it to "table scan" the table for each lookup. When I let the job run over night, it eventually processed some 1,200 rows before I stopped it. If this is your case I found two possible solutions: 1) Add a proper index to the table. 2) Read all the pertinent data from the table into a hash file and do the lookup on the hash file instead of the database table.
Good Luck,
Tony
Just wanted to pass along my findings with a similar situation... I had this exact situation and I found that I was doing a lookup on an external database (Oracle) and didn't give it enough fields to use the correct index, causing it to "table scan" the table for each lookup. When I let the job run over night, it eventually processed some 1,200 rows before I stopped it. If this is your case I found two possible solutions: 1) Add a proper index to the table. 2) Read all the pertinent data from the table into a hash file and do the lookup on the hash file instead of the database table.
Good Luck,
Tony