| Ooops, trying to get used to the Web interface to notes...
Hello,
I had an RZ26L fail on a dual-redundant pair of HSJ40's over the
weekend. I deleted the Dnnn unit ok, but when I tried to delete the
DISKnnn container prior to physically removing the drive, the CLI
on the preferred HSJ hung up (both local and set host). The other
HSJ's CLI would not respond either. I went ahead and pulled the bad
disk from the shelf and inserted a spare disk, hoping that the delete
disknnn command would complete, but hasn't so far. Both HSJ's reset
leds are flashing (running), but I can't prefer the disks to the other
controller, set nowriteback or any other command at this point. The
config is HSOF v2.7J-1, nothing but RZ26L-VA's configured as JBOD's.
My question is, what happens to any cached data if I simply reset the
comtroller to which the bad disk is preferred? The disks preferred to
this controller should failover to the other controller, but will the
data in writeback cache be picked up by the other controller, or what?
I have always been curious as to whether or not the other controller
can access this controller's cache in the event this controller fails.
Thanks, any insight would be highly appreciated,
Dave Robert
615-872-6125 (office)
615-923-8425 (pager)
[Posted by WWW Notes gateway]
|