 |
 |
http://dev.starlink.ac.uk/bugzilla/show_bug.cgi?id=26 ------- Additional Comments From [log in to unmask] 2004-12-08 18:49 ------- Okay Mark. I think I understand. The HDS/NDF system uses a single temp file per process (!). CCDPACK keeps on asking for temp space from HDS. When an HDS file erases a structure the file size does not shrink. Each new temp allocation therefore extends the size of the file despite the file have only one structure. Eventually the file grows too large for HDS to handle. The fundamental bug is that monoliths do not trigger a new HDS temp file each time an action is invoked. Any chance that the ADAM fixed part can re-init the temp file, or force HDS to remove the temp file (which currently is removed during HDS shutdown). Waiting for HDS64 to appear is not tenable since this bug will just trigger again a few steps down the pipeline processing as more calls to the ccdpack monolith cause more temp space to be used. Brian? ------- You are receiving this mail because: ------- You are the QA contact for the bug, or are watching the QA contact.