Assistance required to resolve S001 ABEND

Gaurav_193 · Posted: Wed Mar 14, 2007 9:22 pm

Hi,

I'm facing an issue of bad read on a tape file and the Job abends with S001.

These are the current error messages from spool

IOS000I 035A,BA,SEQ,02,0600,,**,733316,MYJOB 030
.0850214140A1C7200500(3737373737360040)4C06(002C0000)CE0E242574720000

IEC020I DCB EROPT=ABE OR AN INVALID CODE, AND/OR NO SYNAD EXIT SPECIFIED

CEE3250C The system or user abend S001 R=00000004 was issued.

The program preparing the file has file description as
FD FILE1
RECORDING MODE IS V
BLOCK CONTAINS 0 RECORDS
LABEL RECORDS ARE STANDARD.

and the program reading the file has file description as
FD FILE1
RECORDING MODE IS V
BLOCK CONTAINS 0 RECORDS
LABEL RECORDS ARE STANDARD.
01 RECORD1 PIC X(22).
01 RECORD2 PIC X(4118).

The DCB parameters of the file are kept as DCB=(RECFM=VB,LRECL=4122) while it is created in JCL1 and it is accessed while read in a SHR mode in JCL2.

This issue doesn't occur everytime a tape is read but is persistant.

Can anyone explain here what can be the cause of this issue. Is the File description correct while the read program is accessing it?

Thank you in advance!

Regards,
Gaurav

dick scherrer · Posted: Wed Mar 14, 2007 11:55 pm

Hello,

Please look thru the output for something that looks like: S001-04 or -05. There should be an IEC020I message. Pleae post all of the messages from the jes log and other sysout datasets.

If you re-run the job with the same tape does it run or fail again?

One way to get a S001 error is to issue a read after end of file.

Shobana Bhaskar · Posted: Thu Mar 15, 2007 1:57 pm

Yes, Please mention specifically.. Anyhow see any of the below can be the reason.
There can be more than one reason for S001 but all related to dataset length/blksize
S001-0 THE LRECL AND BLKSIZE SPECIFIED IN DCB DIFFERS FROM LRECL AND BLKSIZE INDICATED IN DATA SET.
S001-1 RECFM=U WAS SPECIFIED, BUT NO LRECL WAS SPECIFIED OR -
A WRITE WAS ATTEMPTED TO A DATA SET THAT WAS OPENED FOR INPUT OR -
READ WAS ATTEMPTED TO A DATA SET THAT WAS OPENED FOR OUTPUT OR -
INCORRECT CONCATENATION (I.E. LARGEST BLKSIZE NOT FIRST)
S001-4 THE MOST PROBABLY CAUSES OF A S001-4 ABEND ARE:
I/O ERROR BECAUSE OF DEFECTIVE VOLUME OR DEVICE OR -
AN EMPTY DATA SET OR -
OUT OF SPACE

Gaurav_193 · Posted: Thu Mar 15, 2007 8:58 pm

Thank You Dick and Shobana for the inputs!

Here are some clarifications,

1) On restarting the Job from abended step, it completes sucessfully and that is the big question why and how if this is related to a bad read from tape?
The job is reading the Key from instream data in JCL and retrieving the Key record match from Tape File to VSAM file.
The Tape file is opened in Input mode and VSAM in I-O mode.
Also, can you please explain if this can be a space issue as suggested.

The error that it gives is coded as SEQ under IOS000I message which means
SEQ Block identifier sequence error.
Do you know of what exactly this means?

I'm pasting the spool messages again for reference

IOS000I 035A,BA,SEQ,02,0600,,**,733316,MYJOB 030
.0850214140A1C7200500(3737373737360040)4C06(002C0000)CE0E242
IEC020I 001-4,MYJOB,STEP5,FILEI,035A,733316,
IEC020I FILE.ON.TAPE

IEA995I SYMPTOM DUMP OUTPUT 034
USER COMPLETION CODE=4039 REASON CODE=00000000
TIME=20.24.35 SEQ=05474 CPU=0000 ASID=0095
PSW AT TIME OF ERROR 078D1000 88BE843E ILC 2 INTC 0D
NO ACTIVE MODULE FOUND
NAME=UNKNOWN
DATA AT PSW 08BE8438 - 00181610 0A0D58D0 D00498EC
AR/GR 0: 80CBB81E/84000000 1: 00000000/84000FC7
2: 00000000/0002E250 3: 00000000/0002000D
4: 00000000/08BDD258 5: 00000000/0002DD9C
6: 00000000/00000000 7: 00000000/3FD368B8
8: 00000000/0002E250 9: 00000000/0002F90E
A: 00000000/0002DD9C B: 00000000/08BE8368
C: 00000000/00027A88 D: 00000000/00030A50
E: 00000000/88BDC3A0 F: 00000000/00000000
END OF SYMPTOM DUMP

Thanks again for your co-operation and interest in resolving this issue!

Regards,
Gaurav

dick scherrer · Posted: Thu Mar 15, 2007 10:24 pm

Hello,

You might want to talk with the Operations people about the possibility that the tape is dirty/damaged or that there is a tape drive problem.

If this persists, it is likely that one or the other needs cleaning or repair. If the tape is permanantly damages, you will need to look for a way to re-create the tape.

If this is an old tape, you might try to copy it to a new volume. Many sites have a policy that data that is to be kept for long periods have a backup and are copied to 2 new volumes each year.

Does this happen with multiple tapes or just this one?

Suryanarayana.tadala · New User Joined: 03 Nov 2005 Posts: 43 Location: St.Louis

Looks like solution provided by Dick is apt. In most of the cases there could be a problem at the operations during their maintanace.

Let us know if there is any other possibilty for this type of problem

Gaurav_193 · Posted: Fri Mar 16, 2007 5:18 pm

Hi Dick,

Actually we archive certain records periodically from production files to tapes and when some record say for some reference is required we pull it out from tapes and put back in files.
Everytime the archival records are written on to different tape versions. This ABEND was not persistent in past but had started occuring only recently.

The point is how the job completes sucessfully on restart from abended step if we have a bad tape volume or device problem.
So i should answer your query here is this happens with multiple tapes.

I will talk with operations team and and have them look on to the condition of tapes.

I will update you as and when i get this resolved.

Thank you Dick and Suryanarayana for your assistance!

Regards,
Gaurav

dick scherrer · Posted: Fri Mar 16, 2007 7:16 pm

You're welcome

If you have several abends and can view the output from them, it would be a good idea to check and see if the abends all occur on the same tape drive or in the same silo (if your site uses silos).

It would also be a good idea to see if the problem tapes were all put in use about the same time (in the same shipment). Sometimes tapes are defective and this is only found after some amount of time or some number of uses. If they are unable to initialize, this is detected before any data might be lost. If they degrade in unacceptable fashion (they will all have to be replaced eventually), that is harder to isolate.

Good luck and let us know if we may offer suggestions