To my astonishment, the problem has turned out to be a bad power supply after all. I didn't believe it until some of the SCSI drives in the same box started acting up also.
Score: ZFS: 1, $180 OCZ PowerStream 600: 0 :) I would still argue that a salvager would be a useful addition to ZFS for situations like these. While I managed to back up most of the most valuable data before I lost the pool, as it turned out there were a couple of recently-modified files I didn't have copies of. I wound up writing a C program to search the disks for strings that I knew were in these file; for each match, reading in the disk contents in the vicinity with `dd'; and poring over the data with Emacs to see if this was the version of the file I needed and if so, to pull out the contents. This mostly worked, but was tedious. Given the amount of redundancy in a ZFS pool, it seems like a salvager should be able to recover a lot of stuff (not that it will be easy to write one). This message posted from opensolaris.org _______________________________________________ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
