CA 1 14.0 : Batch jobs error IEFTMS70 2YY-104 OPCODE=36 with Abend S2FF

Document ID : KB000101177
Last Modified Date : 14/06/2018
Show Technical Document Details
Issue:
CA 1 14.0 .

When the CA 1 Tape Management Catalog (TMC) is initially allocated and formatted, one of the requested input parms to provide is the number of ALLOCATED DSNBs (Data Set Name Block)  and this will affect the TMC space allocation.

In fact this is the formula we suggest to calculate it: 

   Cylinders = (3 + V + (D / 2)) / RC

- V
   The number of volume records
- D
   The number of DSNB records
- RC
   The records per cylinder for device type and blocking factor
All these values should be rounded up.

Sometimes it can happen that batch jobs abnormally end with CODE=S2FF and  the following error messages : 

09.36.09 J0109433  IEC205I SYS00001,SCY2ABMR,P030,FILESEQ=1, COMPLETE VOLUME LIST,  956
    956             DSN=X400DB2Y.IC0O0.DB0Y.BMCACM.BMCHIST.GPNKJDU6,VOLS=TA8925,
    956             TOTALBLOCKS=1
 09.36.09 J0109433  IEFTMS70 2YY-104 OPCODE=36
 09.36.09 J0109433  IEFTMS70 ***** CA 1 ABEND,IO,36 *****                  **
 09.36.10 J0109433  IEF234E K A9DE,TA8925,PVT,SCY2ABMR,P030
 09.36.10 J0109433  TMS014  IEF234E K A9DE,TA8925,PVT,SCY2ABMR,P030
 09.36.10 J0109433  IEA995I SYMPTOM DUMP OUTPUT  981
    981             SYSTEM COMPLETION CODE=2FF  REASON CODE=00000068
    981              TIME=09.36.09  SEQ=51948  CPU=0000  ASID=0577
    981              PSW AT TIME OF ERROR  070C1000   BEDD65D2  ILC 2  INTC 0D
    981                NO ACTIVE MODULE FOUND
    981                NAME=UNKNOWN
    981                DATA AT PSW  3EDD65CC - 00181610  0A0DD70F  D048D048
    981                AR/GR 0: 00000000/00000000_84000000   1: 00000000/00000000_842FF000
    981                      2: 00000000/00000000_002FF000   3: 00000000/00000000_3EE016C0
    981                      4: 00020003/00000000_00000004   5: 00020003/00000000_009FF7C8
    981                      6: 00000000/00000000_009A0010   7: 00000000/00000000_BEE02AA8
    981                      8: 00000000/00000000_00CE51A4   9: 00000000/00000000_00000000
    981                      A: 00000000/00000000_00CEC000   B: 00000000/00000000_3EE026C0
    981                      C: 00000000/00000000_3EDFC2C8   D: 00000000/00000000_009A0158
    981                      E: 00000000/00000000_00000068   F: 00000000/000001B4_00000068
    981              END OF SYMPTOM DUMP


How is it possible to bypass this error ?
Environment:
Z/OS - CA1 
Resolution:
One of the main reasons why this abend can occur is that the current number of 'USED DSNB' is equal to the number of 'ALLOCATED DSNB' in the TMC and no more DSNBs are available for use, so it is necessary to do some checks in this area and quickly establish if this constraint of DSNB depend on any possible error on the TMC Chains or if the maximum number of DSNB has been reached and  should be increased to allow batch tape activity to progress. 
Here what can be useful to check:

- from the ISPF / CA1 Interface it is possible to immediately verify the DSNB situation under CA1 TMC RECORD panel .
   For Example: 

        DSNB ALLOCATED    1200000         DSNBUSED 1199999 
       (only one FREE DSNB) 

- running the TMSPTRS Utility for Maintain Multivolume, Multi-Data Set Chains, no pointer errors are got, so all the DSNB are available for use

- running a new Housekeeping batch run,  all the expired tapes are again scratch and unused DSNB again available for use 

The parm THDSNB  in TMOOPTxx CA1 Init member, provides the THRESHOLD percentage at which the product should start issuing in Console the warning highlighted message IEFTMS9, alerting on the current utilization of DSNB.   If not specified, the default value for this parm is 90. 

The message looks like the following:

  *IEFTMS9 USED DSNBS HAVE PASSED THE THRESHOLD, CURRENT UTILIZATION IS 99%

and during normal processing, as the utilization climbs, a new IEFTMS9 message is issued showing the new utilization percentage and the previous message is deleted.
This message can be trapped by an Auto-Operator Product (Like CA OPS/MVS or IBM Tivoli) that will manage the urgency sending specify alerts/emails. 

If this message is lost or not managed appropriately and the DSNBs CURRENT UTILIZATION reaches 100%, all subsequent batch jobs using tapes to create new datasets, start abending S2FF with error:

 09.36.09 J0109433  IEFTMS70 2YY-104 OPCODE=36
 09.36.09 J0109433  IEFTMS70 ***** CA 1 ABEND,IO,36 *****          

and it is necessary to run a Dynamic TMC TMSXTEND job increasing the number of ALLOCATED DSNB, to allow batch activity to progress.
As per the TMC space formula, also the TMC new space allocation should be increased. 

If the abend is not solved even after this activity, Please open a case with CA Support.