JMR611I PKRDALIN R/C = 970C0000 - DSN = nnnn.JOBLOG.DAILY.IND23344 in addition a JMR731A message is generated

Products

• JMR

Issue/Introduction

JMR creates new DAILY file and INDEX file but in the process runs out of space with the following error messages:

JMR611I PKRDALIN R/C = 970C0000 - DSN = nnnnn.JOBLOG.DAILY.IND23344         
6060  JMR731A PKRDALIN - RESOURCE UNAVAILABLE, REPLY 'RETRY' OR 'CANCEL'.      
R 6060,RETRY                                                                   
 JMR611I PKRDALIN R/C = 970C0000 - DSN = nnnnn.JOBLOG.DAILY.IND23344         
6061  JMR731A PKRDALIN - RESOURCE UNAVAILABLE, REPLY 'RETRY' OR 'CANCEL'.      

Environment

z/OS any level

JMR 4.6

Cause

Insufficient space allocation for the INDEX or DAILY file

Resolution

INDEX files only contain single line entries. These files should never be full. The only way this is full is if there is a large amount of jobs being spooled and the file cannot add additional lines to the file. 

This file must be contiguous in its allocation.

The correct way to allocate a new INDEX file is to use the "PGM=PKBLDIDX"  program. You can simply create a new INDEX file that is larger and then use IEBGENER to copy the old data into it. However, it may no longer match the DAILY file information. You may have lost a record or more. This is the reason we recommend PGM=PKBLDIDX.

The sample JCL is in "prefix.CAIPROC(CAPKBLDX))"

For input, use the current DAILY file. For output you use a newly created INDEX file that is EMPTY.

The correct DCB attributes for the INDEX file are:

DSORG =PS

RECFM = FB

LRECL = 75

BLKSZE = 27975  (or what is more efficient on your system)

Of course the required space allocation is up to you.

Keep in mind that there is one line for each job entry in the DAILY file. Many jobs going into the DAILY file means a larger INDEX file and a larger HOLD file. 

PKBLDIDX will read in the DAILY file and create new entries for everything successfully spooled from the HOLD file. You will then have a new INDEX file that matches the DAILY file.

The JMR spooling process is as follows:

The job is spooled to the HOLD file and removed from the JES SPOOL. JMR reads the entry in the HOLD file, creates a one line entry in the INDEX file. JMR then compresses the entry and writes it to the DAILY file. The HOLD file entry is then cleared, leaving room for more jobs to be spooled.

If you run out of space in the HOLD file with a B37 (file runs out of space) or a D37 (Disk runs out of space for an extent) then this could cause unexpected results with JMR in regard to the INDEX/DAILY files. It is always a good practice to build a new INDEX from the DAILY file, then allocate a larger HOLD file, and use IEBGENER to copy the old file into the new one. You can then rename everything back to the regular production names.

Startup JMR again and all should be good. It will pick up where it left off. No data will be lost.