Created attachment 413310 [details] cciss_vol_status.cron I've got two servers with the hpsa driver now, so I can actually test this. The cciss_vol_status program will succeed even on drives that don't exist, so we can simplify the drive-finding logic a little bit. The only caveat is that it still prints to stderr for wrong/missing drives, so we have to hide that (and other) stderr output. @grknight and @blueness: I recall you guys have the hardware for this. If not, sorry!
Created attachment 413590 [details] cciss_vol_status.cron-r1 Here's another version that doesn't ignore stderr, but requires a temp file. The only other way I could come up with would be to run the command twice. I also switched the backticks to $() in this one -- apparently the $() is POSIX, just not supported by ancient shells. I tested it with both cciss/hpsa and on a random non-HP desktop machine, but (unfortunately?) I don't have any dead hard drives at the moment. If somebody wants to pull a drive out of a server, great =)
Good news! One of my hard drives died. I can confirm that both versions of the cron job I posted report the degraded array.
Very sorry for the delay on this, this is in the tree now as part of the 1.12 ebuild.
