[Info-vax] SCSI errors

[Kenneth Fairfield]

I was waiting for someone else to spot this, but...

Note the "large" error count is on DSA110: *and* 
on PEA0: on JANDER. There is *not* a large error
count on any of the SCSI disks nor on the SCSI
device driver, PKB0:

My recollection is that PEA0: the "cluster 
communications" device. (Please, someone, help me
out here! My last Alphas were rolled out the door
yesterday...and they've been cold since last 
February, RIP.)  Since shadow sets are a visible
cluster wide, and it is only the shadow set itself
that is logging errors (not its members), I'd say
this points to a cluster communications problem.
More precisely, a network problem (since this is
an NI cluster).

Forget about the SCSI stuff; it's fine.  Look at 
your network interface to JANDER.

    -Ken