We think this is a serious bug ourselves. It happens to more than spares - kinda weird behavior to be sure. That's why we have at least 2 spares per filer, and we jump all over failed disks.
Hmmm. Y'know, now that I think of it, _one_ of the disk failures we had recently was like this too - I was thinking that it might have been introduced in ONTAP 6.x, but I'm still running 5.3.7R3. Now it seems more likely that it's a disk firmware issue, since the few disk failures we had up until the "disappearing drive" were all garden variety bad blocks and such. The new symptom only appeared after I bumped the ONTAP rev, added some new drives, and rolled the drive firmware to current levels.
I'm sure with all the autosupport feedback that NetApp can piece together the pattern here... but trying to reproduce and isolate a problem like this and find a fix for it doesn't sound like any fun at all. :-)
-- Chris
-- Chris Lamb, Unix Guy MeasureCast, Inc. 503-241-1469 x247 skeezics@measurecast.com