| Title: | + OpenVMS Clusters - The best clusters in the world! + |
| Notice: | This conference is COMPANY CONFIDENTIAL. See #1.3 |
| Moderator: | PROXY::MOORE |
| Created: | Fri Aug 26 1988 |
| Last Modified: | Fri Jun 06 1997 |
| Last Successful Update: | Fri Jun 06 1997 |
| Number of topics: | 5320 |
| Total number of notes: | 23384 |
I have a customer running VMS 5.4-1 on a 3 node VAX 6000 cluster.
Currently his system disk is a RA82 that is shadowed by another
RA82. He wants to move off of these drives and move the system disk
to RA92's but they can't stand much if any down time. So heres what
I was wondering can we remove one of the RA82's from the shadow set
and then bring one of the RA92's into that shadow set and let the system
do a shadow copy. Then after the shadow copy has completed then drop
out the remaining RA82 and add another RA92 into the shadowset.
The only other way that I know we could do this is shutdown the cluster
do a stand-alone back of the system disk then copy this to the RA92
and then do a shadow copy to the other RA92. These guys want to do
it the safest way with the least amount of downtime.
| T.R | Title | User | Personal Name | Date | Lines |
|---|---|---|---|---|---|
| 5217.1 | NO - same geometry required | HAN::HALLE | Volker Halle MCS @HAO DTN 863-5216 | Fri Jan 24 1997 12:40 | 11 |
No, you can only shadow disks, which have the same no. of blocks and
the same geometry.
In this case, running a BACKUP/IMAGE/IGNORE=INTERLOCK from the 'old' to
the 'new' system disk might work, if the contents of 'open' files on the
'old' system disk is not important. For minimum downtime, you might
then just 'minimum' boot from the new system disk and manually copy those
files, that had been opened for write during the backup.
Volker.
| |||||
| 5217.2 | AMCFAC::RABAHY | dtn 471-5160, outside 1-810-347-5160 | Fri Jan 24 1997 18:19 | 12 | |
V5.4-1??? Geesh, V7.1 has shipped, any chance of upgrading? Have they bought a prior version support contract yet? Is V5.4-1 support even available at all? In general this is not a prefectly strong point in OpenVMS yet, let alone in such an old version. As suggested in .1, use the cluster to good effect; create a second system disk on the newer (and I use the term loosely) technology. Only one node need be 'down' during this. It is basically like a rolling upgrade. Once the new system disk is done then the old technology can be retired. Can the application run on only two of the three nodes? | |||||