Mark Hahn wrote:
I have shuffled the cards,cables and physical drives around to determine
that this is not a problem with any of them individually - no matter the
combination, it only ever happens to drives that are at sd[abcd] (ie: if
I rejig the hardware so the drive at /dev/sdh, which was working fine,
is on a different cable and controller, but appears at /dev/sdb, it will
produce the errors).
did you test the case where all disks had power, but only 8 were plugged
into controllers?
no individual card, cable or drive was responsible. The errors _only_
occur with three cards in the system, _only_ with whichever drives are
attached to the "first" controller (ie: sd[abcd]) and _regardless_ of
other system activity.
the "first" card would correspond to position on the PCI bus (slot),
so perhaps that card is getting iffy power. but did you actually move
around which power cables are supplying which disks?
Power supply was also one of my suspicions, so I tried powering up half
a dozen of the drives off another ATX power supply I had and the
remainder off the system PSU. The same problems occurred, which I think
rules out the possibility of insufficient power (I did try with all
drives powered, but only two cards installed and that worked fine - but
I haven't moved around the power plugs of individual drives).
I'll try again tonight with all the drives powered off their own PSU,
just to be sure.
CS
-
To unsubscribe from this list: send the line "unsubscribe linux-raid" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
