> Hello LAVC gurus, I need your help:
> Given: LAVC consisting of boot nodes A and B and satellites X and Y.
> A and B are vax 11/750s with identical system disks except for
> nodeid, and similar node specific files.
> Satellite X boots from A
> Satellite Y boots from B
> Now, Boot node A goes down. Satellite X is as a result hung and can't
> reboot from A which is down.
> Desired result: Dynamic reconfiguration so when X reboots, B answers the
> call because A doesn't. I.e., X should boot from A, but if A is
> down, boot from B; Y should boot from B, but if B is down, boot from
> A. Is it possible to achieve this result without a system manager
> running cluster_config at the time of failure??
> Is your NETNODES_REMOTE.DAT file in SYS$SPECIFIC? I would expect
>that any eligible boot node should reailize that they can boot from A or B,
>it just depends on which one answers first.
> Make sure that the information for the nodes (MCR NCP SHOW NODE
><nodename> CHAR) matches on both boot nodes, and update those items that
It appears to me that David's reply is incorrect because Todd says satellite X
hangs when 'A' goes down. This implies that A doesn't just service the MOM
request from X when the boot process starts, but it also serves X's system
disk. Todd says he has 2 system disks which are 'identical'. Well, identical
perhaps, but different... If X uses a non-HSC disk which mounts locally to A,
then there is no way around the problem.
Harvey Brydon | SLB DECnet: ASL::BRYDON
Schlumberger Anadrill | Phase V: %RMS-F-RTB, record exceeds 65535 byte limit
Sugar Land, TX | FAX: (713)274-8399
77478 USA | Telex: 1565030001 SINET ASL SLB
Sol 3 | P.O.T.S.: (713)274-8000 x8281 or 274-8281 (d.i.d.)
Real programmers can't spell "keesh"