|
View previous topic :: View next topic
|
| Author |
Message |
M Lee Klein
New User
Joined: 08 Feb 2022
Posts: 39
Location: USA
|
|
|
|
I received a D37 when trying to format 2 new volumes for a dataset (which I uncatalogued). I used Space = (0,3337) for a device 3390 and the new packs have 3337 cylinders or (50040 tracks) available. The other volumes for this dataset have 3337 cylinders except for 3 volumes which have 25500 tracks, 39000 tracks, and 38040 tracks.
Why is the system having a problem with primary space?
MLee |
|
| Back to top |
|
 |
sergeyken
Senior Member
Joined: 29 Apr 2008
Posts: 2018
Location: USA
|
|
|
|
The ABEND code D37 has multiple subcodes (after dash). There is a specific explanation in IBM manuals for each specific reason of D37-xx. Please verify this, rather than asking us to guess what might be wrong.
In many cases there is a text message in ISPF log regarding the reason of ABEND. Try to collect all available information before start guessing. |
|
| Back to top |
|
|
Joerg.Findeisen
Senior Member
Joined: 15 Aug 2015
Posts: 1246
Location: Bamberg, Germany
|
|
|
|
| Also check the DADSM Code and please provide more details about the DSN, and the structure. What JCL/IDCAMS was used to create the DSN and so on. |
|
| Back to top |
|
|
Pete Wilson
Active Member
Joined: 31 Dec 2009
Posts: 580
Location: London
|
|
|
|
You say you coded Space = (0,3337), which I assume you meant as SPACE=(CYL,(0,3337)). That is requesting 0 Primary cyl's space and 3337 cyl's Secondary space, which is a strange thing to do. Is there some reason for that?
You also say you're trying to 'format 2 new volumes' which is strange terminology to use. Are you just trying to add some extra volumes to an existing dataset so it can extend? If so, why would you uncatalog it? If it is cataloged, then potentially you could simply do an IDCAMS ALTER ADDVOLUMES to provide the extra volumes for the dataset to extend to.
It would be a lot simpler if it was SMS managed and had a DATACLAS that would make it automatically extend as required.
It would be interesting to see the dataset information to show how it is really defined. |
|
| Back to top |
|
|
Joerg.Findeisen
Senior Member
Joined: 15 Aug 2015
Posts: 1246
Location: Bamberg, Germany
|
|
|
|
| Pete Wilson wrote: |
| You say you coded Space = (0,3337), which I assume you meant as SPACE=(CYL,(0,3337)). That is requesting 0 Primary cyl's space and 3337 cyl's Secondary space, which is a strange thing to do. Is there some reason for that? |
It makes no sense in this context to EXTEND something. Best use case for this is perhaps DFSORT SORTWK to let it only allocate when SORTWK space is really needed. |
|
| Back to top |
|
|
sergeyken
Senior Member
Joined: 29 Apr 2008
Posts: 2018
Location: USA
|
|
|
|
After SPACE=(...(0)) is processed, there is only DSCB created in VTOC, and nothing else.
I'm afraid the DS extension routine may get crazy while trying to handle SPACE=(...(0,NNNNN)), and not even one track has been allocated yet... 
Run your test with SPACE=(CYL,(1,NNNNN)) to see the difference. |
|
| Back to top |
|
|
M Lee Klein
New User
Joined: 08 Feb 2022
Posts: 39
Location: USA
|
|
|
|
The error was SYSTEM COMPLETION CODE=D37 REASON CODE=00000004
535 TIME=09.30.53 SEQ=00049 CPU=0000 ASID=005
It was an existing dataset where data that needed to be retained was on the other volumes. I was using an Adabas utility to format the dataset after uncataloging. The utility formats the last two volumes and catalogues the dataset with the volumes.
M Lee |
|
| Back to top |
|
|
sergeyken
Senior Member
Joined: 29 Apr 2008
Posts: 2018
Location: USA
|
|
|
|
It remains unclear: what is the relation between a dataset create/delete/catalog/uncatalog, and a volume formatting???
If any volume has been formatted, there is no trace of previous datasets left!
This process is unrelated to catalog/uncatalog a dataset, too. |
|
| Back to top |
|
|
M Lee Klein
New User
Joined: 08 Feb 2022
Posts: 39
Location: USA
|
|
|
|
The ADAFRM Adabas utility prepares the new volumes to receive data when the existing dataset gets too small. The ADAFRM utility catalogs the dataset after formatting the new volumes.
This formatting of the new volumes by the Adabas utility is different from the formatting by an IBM utility like ICKDSF.
But the error received is D37, subcode 0004.
I've contacted the Adabas vendor. I was hoping someone here could help me understand why the error occurred. I thought it might be related to the dataset volumes being of different sizes.
M Lee |
|
| Back to top |
|
|
Pete Wilson
Active Member
Joined: 31 Dec 2009
Posts: 580
Location: London
|
|
|
|
| I expect the utility is only formatting the database dataset, not the volume itself. |
|
| Back to top |
|
|
Pete Wilson
Active Member
Joined: 31 Dec 2009
Posts: 580
Location: London
|
|
| Back to top |
|
|
|
|
