Thanks for the suggestion from Alan, which didn't solve the problem. The
solution was to reinstall the OS, subsets, etc., and re-configure. I think
the configuring, de-configuring both hardware and software, multiple times,
just confused the cluster. Thankfully, it was a development machine, so I
had the luxury of starting from scratch.
Sorry for the late summary.
Wendy
Original Question:
>
>I have two 4100s with DU4.0d-p3, memory channel, HSZ50, OPS, LSM, and PSE
>running. I only have one ase service and two machines.
>
>"dd" doesn't work on the second machine until I move the service. Oddly,
>when the second node is managing the service, I can "dd" to the drd disks
>from both machines and Oracle Parallel Server works just fine. When the
>service is managed on the first machine, OPS won't come up on the second
>machine because it can't read the control file on the drd cluster. The
>service was configured on the first machine.
>
>>From what I have read, I don't have to run
># drd_mknod -f drd[1-6]
>or,
># drd_ivp -r
>on the second machine because the two machines are in the same ase service.
> Or, am I reading the man pages badly? Is this the reason I can't "dd", or
>does someone know what the real problem is?
>
Thanks to Alan for this reply:
Check the error log for reservation conflict messages. The
system where it works probably has the underlying disk
reserved, to prevent access by the other host.
=====================
Wendy Fong
wfong_at_synacom.com
408.296.0303
UNIX is very user friendly,
it's just very particular about
who it makes friends with.
=====================
Received on Mon Aug 30 1999 - 18:25:07 NZST