| Title: | VAX and Alpha VMS |
| Notice: | This is a new VMSnotes, please read note 2.1 |
| Moderator: | VAXAXP::BERNARDO |
| Created: | Wed Jan 22 1997 |
| Last Modified: | Fri Jun 06 1997 |
| Last Successful Update: | Fri Jun 06 1997 |
| Number of topics: | 703 |
| Total number of notes: | 3722 |
A major customer currently has a 3 disk shadow set over two sites. Both
sites are OpenVMS 6.2 H3.
In the main site they have two disks shadowed via host based shadowing
through CI to 2xHSJ40s (one disk on each HSJ40). The third member is at
a remote site but also on a HSJ40. All disks are therefore HSJ40 based
and all are currently configured as JBOD (from HSJ40 perspective).
The first question posed is if the two main site disks are moved to one
HSJ40 and made hardware RAID shadowed does the remote disk (JBOD) need
to be forced into a single disk shadow set for compatibility with the
main site? (Yes all disks will be of the same size).
The second question is, given they do a backup for safety sake, will
they be required to restore any of the disks (either local or remote
site) after the relevant HSJ40s are reconfigured?
The third question is if the 2xHSJ40s (in a HSJ42) are both connected to
each of the the disks in the RAID set ie. multi-pathing of the RAID
set, will the failure of either HSJ40 cause any likely problems in
availability of the RAID set (assume HSOF 3.1, writeback cache etc)
from OpenVMS perspective?
Any related experience would be most welcome.
Wayne
Note: FYI the customer has two VAX 7800s on the main site and one at the
remote conncected via fibre. Also they have 3xHJS42s, 2 at the main
site and one at the remote. The configuration used above was
simplified to minimise the context for the questions. This paragraph
was added to minimise the likelihood of getting a batch of "why do/don't
they" & "have you thought about" questions in reply to this note.
| T.R | Title | User | Personal Name | Date | Lines |
|---|---|---|---|---|---|
| 578.1 | GODOWN::MARSHALL | PaulG.Marshall, Systems Consultant, SPR_CSC | Sun May 11 1997 19:04 | 11 | |
Wayne
I have been following up the first point and talking with the customer,
call me off line please.
Looks like the first point may be possible but I'm unsure whether the
customer will see any benefit...
Regards
Paul
| |||||
| 578.2 | AMCFAC::RABAHY | dtn 471-5160, outside 1-810-347-5160 | Mon May 12 1997 12:48 | 14 | |
re .0:
Good questions. The VMS shadow driver will only care about geometry -- if the
geometry remains the same when you form the controller-based RAID 1 then you'll
be ok. You need to ask the storage folks if they change the geometry presented
to the OS when they form such sets.
From the SHOW DEVICE /FULL command look for the following section;
Total blocks 832527 Sectors per track 62
Total cylinders 1492 Tracks per cylinder 9
Your third question depends on how the controller fails. If it starts a fire
whilst failing that eventually takes out the other controller, well, then ...
| |||||