Client received a dump yesterday for CAS5121E CCI CONV error - RC= 08 RS= 22 ERC= 00000000. I found document ID KB000044996 which mentions this is usually due to a fragmented database. However, if I look at our parms, we do have Y for REORG in the LOGGER statement and the CCILGR proc has SYSUT1.
The dump shows that one of the tasks is taking an extremely long time completing an I/O to the database. Based on the dump 6500 records for 36 systems have been processed. This will result in the original 6500 records being written and deleted from the database along with the responses from each system(6500*36=234000). That total is about 240,500 records written and deleted. This would definitely result in many CI/CA splits. If they are normally processing over 6000 CPF requests per day, per system, I would recommend recycling CCILGR daily, rather than every 2 days to ensure proper I/O performance.
Recycle CCILGR daily, rather than every 2 days to ensure proper I/O performance.