jac@zabriskie.berkeley.edu (Gordon Jacobs) (06/01/90)
I have set up automatic nightly lev 9 dumps to a certain partition on my DECstation 3100 system using the commands: rm -f /dump/filename /bin/dump 9f /dump/filename /partition >& /dump/dump.log The program faithfull executes but sometimes the file written has a size that varies drastically from the amount of information read. For example, in the log, the output of dump will be DUMP: Estimated 15360 bytes output to file named /dump/tools/tools.lev9.Fri DUMP: Dumping (Pass III) [directories] DUMP: Dumping (Pass IV) [regular files] DUMP: 4461568 bytes were dumped to file /dump/tools/tools.lev9.Fri DUMP: Dump is done Trying to read the file with restore -if causes a segmentation violation. This does not happen all the time, but it will occurr reliably after a certain tape dump at a lower level. Any ideas what is wrong? Thanks a lot, Gordon Jacobs
alan@shodha.dec.com ( Alan's Home for Wayward Notes File.) (06/02/90)
In article <25360@pasteur.Berkeley.EDU>, jac@zabriskie.berkeley.edu (Gordon Jacobs) writes: } } [ Example of using automagick backups... ] } } The program faithfull executes but sometimes the file written } has a size that varies drastically from the amount of } information read. For example, in the log, the output of dump will be } } DUMP: Estimated 15360 bytes output to file named /dump/tools/tools.lev9.Fri } DUMP: Dumping (Pass III) [directories] } DUMP: Dumping (Pass IV) [regular files] } DUMP: 4461568 bytes were dumped to file /dump/tools/tools.lev9.Fri } DUMP: Dump is done } } Trying to read the file with restore -if causes a segmentation violation. } This does not happen all the time, but it will occurr reliably after } a certain tape dump at a lower level. } } Any ideas what is wrong? There is a quirk (bug?) in dump that it writes a fairly large dump file when there were no files to be backed up. The size of the file seems to depends on the depth of the directory structure of the file system. When restore sees this (large) empty incremental it dies with the segmentation violation mentioned. I submitted a QAR during the V4 field test and was told it would be "Fixed in a future release." } Thanks a lot, } Gordon Jacobs -- Alan Rollow alan@nabeth.enet.dec.com