We've had an application that has abended a few times in the last week due to storage shortages. Storage consumption begins at a reasonable level, but slowly creeps up over several days until ultimately failing. After collecting a getmain / freemain trace and dump of the task, we were able to isolate the problem to a leak coming from SAFOEUMP / SAFOEUSP at RMID R095981. The trace shows a getmain at offset x'F82' for x11A0' bytes, and a nearly immediate freemain at x'11D4' for x'1198' bytes (all but the first 8 of the getmained area). Because the storage is in LSQA subpool 255, key 0, the page is fixed and the storage creeps. We weren't seeing this until applying recent maintenance to the modules in question.