> Ok, that's a point but would that explain that SCSI cards (like the
> adaptec 78xx AHA-2940 u/w SCSI) would start to have some
> strange behaviors like SCSI timeouts whatever you are trying to do ?
This has pretty much always been my experience with Adaptec controllers,
from Linux .98 to current.
Every one I've had fail has failed flakey like this. Start out with a
bus, maybe five devices on the bus, everything has run OK for a couple of
months, and then I start seeing SCSI errors.
So I mess around, try to eliminate devices to no avail; and finally I swap
out the passive terminator with an active.. Then things run OK again, for a
couple of months. Then errors start creeping in again, timeouts, occasional
bus hangs. Then I swap out the active for a forced-perfect terminator; and
again, things behave for a while.
Then after a couple of months things start flaking out again; and I go
through and replace all the cabling with the shortest cables I can get away
with, and things seem to stablize for a while, then the errors start occuring
again.
And I can't get them to stop for anything. Replace the controller, things
are good for a while, 4 or 5 months later they start flaking again. Replace
the controller again, another 4 or 5 quiet months.
Finally, replace Adaptec with Buslogics (or whatever the name of the
company that bought them)... No more problems.
I don't know what it is about Adaptec's, maybe I've just been unlucky, but
I am personally convinced there is something about the bus drivers or
tranceivers that gradually degrade.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/