View previous topic :: View next topic |
Author |
Message |
s000p Beginner
Joined: 22 Jun 2005 Posts: 12 Topics: 5
|
Posted: Thu Jun 23, 2005 11:57 am Post subject: Abend code U0273 while restarting |
|
|
Hi,
I am facing an abend of U0273 when I restart the job after relisting the Input GSAM file. My program reads in a GSAM file and it is interfaced with some external device through some API's. It got abend due to some problem in that device. I deleted some of the records from the input GSAM file and then restarted the job again.
I observed that sometimes the program run fine, program reads the next record after the last checkpoint and processed normally but sometimes it abend at the very first attempt of reading the GSAM file with error code of U0273.
I cannot understand the arbitrary occurance of this abend. Can anyone put some more light on the cause of this problem.
Thanks,
SP |
|
Back to top |
|
 |
Bithead Advanced

Joined: 03 Jan 2003 Posts: 550 Topics: 23 Location: Michigan, USA
|
Posted: Thu Jun 23, 2005 12:40 pm Post subject: |
|
|
The restart uses the RBA (relative byte address) from the start of the file as a starting place. If you edit the dataset, you could change the address of the restart record, especially if the dataset contains short-blocks. |
|
Back to top |
|
 |
s000p Beginner
Joined: 22 Jun 2005 Posts: 12 Topics: 5
|
Posted: Thu Jun 23, 2005 11:16 pm Post subject: |
|
|
OK..
Can you explain what is short-block and will it make a difference if we have long blocks??
Why the occurance is in random fashion and not consistently we get this abend whenever we modify/delete the records in the GSAM file?
Thanks,
sp |
|
Back to top |
|
 |
Bithead Advanced

Joined: 03 Jan 2003 Posts: 550 Topics: 23 Location: Michigan, USA
|
Posted: Fri Jun 24, 2005 8:33 am Post subject: |
|
|
Was the file generated by another IMS program? |
|
Back to top |
|
 |
s000p Beginner
Joined: 22 Jun 2005 Posts: 12 Topics: 5
|
Posted: Fri Jun 24, 2005 12:52 pm Post subject: |
|
|
No, This file is created manually by me for testing purpose only. |
|
Back to top |
|
 |
Bithead Advanced

Joined: 03 Jan 2003 Posts: 550 Topics: 23 Location: Michigan, USA
|
Posted: Fri Jun 24, 2005 1:10 pm Post subject: |
|
|
Short blocks will not be a problem in this case. Use of Short blocks allows IMS to force data be written to DASD rather than remain in the buffer until a "regular sized" block is written. This means that there is a valid restart point in event of I/O failure.
Are you possibly deleting records that had already been processed before the last ckeckpoint? |
|
Back to top |
|
 |
s000p Beginner
Joined: 22 Jun 2005 Posts: 12 Topics: 5
|
Posted: Sun Jun 26, 2005 11:01 pm Post subject: |
|
|
Hi,
Yes I am deleting the records from the file before the last checkpoint.
Please tell me from where I can see that a particular GSAM is a short block or a long block.
Thanks,
sp |
|
Back to top |
|
 |
Bithead Advanced

Joined: 03 Jan 2003 Posts: 550 Topics: 23 Location: Michigan, USA
|
Posted: Tue Jun 28, 2005 7:44 am Post subject: |
|
|
You must not delete records that are before the last checkpoint. These have already been processed and any updates have been committed. The restart will postion IMS at a given number of bytes from the start of the file. Deleting records will cause the restart to be positioned incorrectly.
You cannot see short blocks, you just have to be aware of them. They are created if the file is created by IMS as a GSAM and has updates and takes checkpoints. |
|
Back to top |
|
 |
vikaspaniker79 Beginner

Joined: 01 Aug 2005 Posts: 12 Topics: 4 Location: India
|
Posted: Fri Aug 05, 2005 6:43 am Post subject: |
|
|
Hi,
Can you also check if you have the right key in the checkpointed area... u mentioned that it is working sometimes fine ...maybe at some logical record the uniqueness of the checkpointed value is lost .... _________________ Cheers,
Vikas |
|
Back to top |
|
 |
|
|