Archive trigger process is not using the PRIMARY and SECONDARY values specified in the Archive Application parameters / global variable values.
Parm Name Name Parm Value LOADED for execution
===========================================================
BASENAME ARCHIVE GLOBAL0.ARCH_TRACK.
GDGMODEL ARCHIVE HLQ.OPSLOG.ARCHPTRN
PRIMARY ARCHIVE 2000
SECONDARY ARCHIVE 1000
SMS ARCHIVE YES
STORCLAS ARCHIVE
VOLSER ARCHIVE (PARM NOT LOADED)
Global Prefix: GLOBAL8.Z9_ARCHIVE
Subnode Name Nodes Subnode Value
BASENAME 0 GLOBAL0.ARCH_TRACK.
GDGMODEL 0 QAUP1E.OPSLOG.ARCHPTRN
PRIMARY 0 2000
SECONDARY 0 1000
SMS 0 YES
STORCLAS 0
VOLSER 0
Archive parms:
Name Value Description Row 3 of 9
ARCHIVEHLQ 'HLQ.OPSLOG.ARCH OPSLOG ARCHIVE HIGH LEVEL QUALIFIER
ARCHIVEMODEL 'HLQ.OPSLOG.ARCH OPSLOG ARCHIVE MODEL DATASET
ARCHIVEPROC 'OPSARCH' OPSLOG ARCHIVE PROC
ARCHIVEUNIT '' OPSLOG ARCHIVE DEFAULT UNIT
ARCHIVESTORCLAS '' OPSLOG ARCHIVE STORAGE CLASS
ARCHIVEVOLSER '' OPSLOG ARCHIVE VOLUME SERIAL
ARCHIVETRIGGER '250000' MESSAGE COUNT OR TOD AT WHICH TO TRIG
Archive execution showing Primary and Secondary = 223,223 instead of 2000,1000
------ JES2 JOB STATISTICS ------
03 MAY 2021 JOB EXECUTION DATE
2 CARDS READ
133 SYSOUT PRINT RECORDS
0 SYSOUT PUNCH RECORDS
12 SYSOUT SPOOL KBYTES
0.13 MINUTES EXECUTION TIME
1 //OPSARCH JOB MSGLEVEL=1
2 //STARTING EXEC OPSARCH,SSID=OPSS,LOGNAME='OPSLOG',TRACK=(223,223)
03MAY 15:47:01 OPS4403O 250002 messages since the last archive. Current message number is 60763831 - OPSLOG
03MAY 15:47:02 $HASP100 OPSARCH ON STCINRDR
03MAY 15:47:02 $HASP373 OPSARCH STARTED
03MAY 15:47:02 IEF403I OPSARCH - STARTED - TIME=15.47.02
03MAY 15:47:09 SVM4000I OPSARCH, OPSARCH, SYS00001, ATTEMPTING SPACVOLA FOR HLQ.OPSLOG.ARCHIVE.G0550V00, 1EG
03MAY 15:47:09 IEC030I B37-04,IFG0554A,OPSARCH,OPSARCH,SYS00001,B009,1EG004,
03MAY 15:47:09 QAUP1E.OPSLOG.ARCHIVE.G0550V00
03MAY 15:47:09 IEA989I SLIP TRAP ID=ABCD MATCHED. JOBNAME=OPSARCH , ASID=009C.
OPS/MVS
The new OPSLOG archive process internally calculates the amount of space in tracks needed to copy the number of OPSLOG records for the current archive request and then passes this amount on the internal start command. The total space calculated for the archive is divided by 16 and that amount is specified as the primary and secondary amounts to avoid a conflict with SMS for storage constraint relief classes and a rejected request if the requested volume/s do not have the entire needed space in a contiguous area.
Look at the dataset that was allocated for the archive request that received the B37 abend. There was not enough space on the volume to handle the request - 223 tracks x 16 extents= 3568 tracks; only 2007 tracks were allocated/available on volume 1EG004. Based on the manual archive execution of the same data, 13 extents is what would have been needed on the failing request to accommodate the archive request. So the total space needed was not available at the time of the request on volume 1EG004.