Hi, There are several possible causes
1. everything just takes a little bit more storage than it used to 2. You have a storage leak 3. You have fragmentation How to report storage usage 1. SDSF real storage gives a bit of an indicator... it should increase then level out. it it increases every day, it may indicate a leak. It is real usage - not virtual storage - so only a hint 2. Check you have region=0m and memlimit=0 so you use all that is available in the region 3. Use the rptstg <https://www.ibm.com/docs/en/zos/2.4.0?topic=ulero-rptstg> https://www.ibm.com/docs/en/zos/2.4.0?topic=ulero-rptstgleoption to display usage, and get recommendations on initial values 4. Consider use of storage pools if you are not using them - it prevents storage fragmentation To see if it is a leak ---- run for a day, and look at the rptstg usage. Run for 2 days, and do the same. etc. They should be similar. If there is increasing storage usage, it might indicate a leak (or it might just indicate more workload) Colin Once it is working you can look into threadheap and thread stack ( 64 bit) which may help On Tue, 24 Dec 2024 at 20:44, Matthew Stitt <mathwst...@bellsouth.net> wrote: > Ever since upgrading to z/OS V3R1 and using JAVA 17, zOS MF starts > normally, but will abend after several days with either an S80A or an S878. > I'm beginning to think it could be Language Environment heap storage > management. I also have gotten JAVA JVM dumps, but cannot make sense out > of them. > > All I do is restart the z/OS MF subsystem. > > I did not have this under z/OS V2R5 with JAVA 8 64 bit. > > Any thoughts? > > Matthew > > ---------------------------------------------------------------------- > For IBM-MAIN subscribe / signoff / archive access instructions, > send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN > ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
