book
Article ID: 191741
calendar_today
Updated On:
Issue/Introduction
This procedure will identify and eliminate tape volume serial numbers found within the EPIC catalog. You can perform this procedure while
EPIC is active in your system (there is no adverse impact on running jobs requesting EPIC services).
Environment
z/VSE any level
CA Epic for VSE 5.1 and above
Resolution
1) IDENTIFY DUPLICATE VOLUME SERIAL NUMBERS
Run the following job step (a label for SORTWK1 is assumed to be
available in the STDLABEL or PARSTD label areas):
// EXEC TSIDUTL
DSPLY ALL,SEQ=SERIAL
/*
If EPIC message EP210 is issued, there is at least one and possibly
multiple duplicate volume serial numbers.
The resulting listing needs to be reviewed very carefully by scanning
down the SERIAL NO column looking for an asterisk (*) just to the right
of the volume serial number displayed in that column. If a volume
serial number is displayed more than once but not flagged with an
asterisk, it only means that more than one file is active on that volume
serial number and is not to be considered a duplicate.
You typically find a volume serial number duplicated only once with an
asterisk appearing to the right of the second occurrence. It is also
possible that a single volume serial number can be repeated several
times with some or even all of the repetitions flagged with an asterisk.
This means that the volume serial number is duplicated several times.
Make a note of any unique volume serial number that is identified as a
duplicate.
2) PHYSICALLY RESERVE ALL IDENTIFIED DUPLICATE VOLUME SERIAL NUMBERS
Take physical possession of each unique volume identified as a duplicate
in Step 1. This ensures that the disposition (either active or scratch)
of the identified volume is correct when you are done.
3) PURGE DUPLICATE VOLUME SERIAL NUMBER(S)
Use either the TSIDMNT batch catalog maintenance program or your
online catalog maintenance transaction to PURGE (not DELETE) each unique
volume serial number identified as a duplicate. If the purge request
fails with EPIC message EP116 VERSION CONTAINS ACTIVE FILES, go back to
the TSIDUTL DSPLY ALL,SEQ=SERIAL listing created in Step 1 and scratch
all dataset versions listed as active for that volume serial number and
then attempt the purge request again.
4) REBUILD THE CATALOG INDEX
Run the following job step (a label for SORTWK1 is required and must be
either a start-track-1 or uncontrolled dataset as documented in the
"EPIC Installation and System Guide" - this sample JCL shows a typical
start-track-1 dataset which needs the disk pool name changed if POOL01
is not appropriate for your installation):
// DLBL SORTWK1,'===.TSIDUTL.INDEX.SORTWK1(D)'
// EXTENT SYS001,POOL01,,,1,15
// EXEC TSIDUTL
INDEX
/*
It is normal for other jobs running on any system which shares this
EPIC catalog to wait during the time the TSIDUTL INDEX function is
processing because the catalog is enqueued for the duration of this job
step.
5) DETERMINE IF ALL OCCURRENCES HAVE BEEN ELIMINATED
Use either the TSIDMNT batch catalog maintenance program or your
online catalog maintenance transaction to display each unique volume
serial number identified in Step 1. For each unique volume serial
number identified in Step 1 that is still found in the catalog, return
to Step 3. Continue with Step 6 only when every unique volume serial
number identified in Step 1 no longer exists in the catalog.
6) DETERMINE TRUE CONTENTS OF RESERVED VOLUMES
To determine the true contents of each volume reserved in Step 2,
process that volume with the DITTO Tape Map function. Here is a
sample of a typical DITTO batch control statement to accomplish the TAPE
Map function:
$$DITTO TMP INPUT=cuu,NFILES=EOV,NBLKS=2
DITTO will display an 80-byte record called a 'HDR1' which identifies
the beginning of another data file found on the volume. Here is a
partial field layout of that 80-byte record:
Start End Length Description
1 4 4 'HDR1'
5 21 17 Dataset name (padded with blanks if less than 17 characters or truncated to 17 characters if longer)
32 35 4 Dataset sequence number (zoned decimal - 0001 indicates a prime dataset, greater than 0001 indicates a subordinate dataset)
42 47 6 Creation date (zoned decimal julian date (yyddd) right justified)
7) REPLACE RESERVED VOLUMES BACK INTO CATALOG
If none of the datasets found on the volume are to be considered active,
use the TSIDMNT batch catalog maintenance program or your online
catalog maintenance transaction to add the volume serial number to the
prime dataset as version 1 and then scratch version 1 of that dataset.
If any dataset found on the volume (either the prime dataset or any
subordinate dataset) is to be considered active, use the TSIDMNT batch
catalog maintenance program or your online catalog maintenance
transaction to add the volume serial number to the prime dataset as
the appropriate version and then retain that version to the appropriate
expiration date. Add the subordinate dataset version(s) still
considered active to its respective dataset name using the appropriate
version number while specifying the file sequence number value as shown
by the HDR1 record and then retain that version to the appropriate
expiration date.