Over the weekend I managed to complete the upgrade and I've summarised the
procedure.
The hardware/software configuration was:
2 * Alpha 4100s running Digital UNIX 4.0D - ECO Patch #3
Shared SCSIs connecting to 2 HSZ50 array controllers
Trucluster ASE 1.5
LSM generated volumes used as raw disks
Objective:
Upgrade HSZ firmware to 5.5 (needed to support the new disks)
Replace 4 Gig drives with 18 Gig drives
Procedure (derived from TruCluster Available Server Software manual):
* There is most probably a good manaual or something that lists this for
you, but I couldn't find it at the time, so a lot of this was guess work.
1) Stop (OFFLINE) all ASE services
2) bring the LSM used disks online:
voldisk online rzXXX rzYYY rzZZZ
3) Import the disk group:
voldg import disk_group
4) Restart the volumes
volrecover -sb disk_group
5) Modify the LSM configuration to delete the volumes and remove the 'to be
replaced' disks from LSM
6) Deport the disk group and offline the remaining disks
7) Shut the system down and make the changes to the HSZ50s. Install new
firmware and install the new disks (we used Command Console 2.0)
8) Bring the system into multi user mode and create the /dev devices
9) Bring the disk_group back online as in steps 2 to 4
10) Modify the LSM configuration to add the new disks and make your new LSM
volumes
11) When the LSM disks are ready to be used, make the ASE changes:
asemgr - Managing ASE Services - Service Configuration - Modify a service -
disk_service on server - General service information - Add a UFS file
system, Advfs fileset, LSM volume, or raw disk - enter the name of the any
new online LSM volume. This should make ASE find the changed disk group
modifications and make the changes for you. It will delete the current
service and re-add it.
12) Because we're using raw disks I deleted the unwanted UFS disk entry.
13) Deport the disk group and offline the remaining disks again
14) Start (ONLINE) the ASE service and populate whatever disks were
modified.
Original Question:
> Hello,
>
> I'm about to undertake (for the first time) an upgrade
> of several hard drives by replacing 4 gig drives with
> 18 gig drives in a 2 * HSZ50, 3 server cluster environment.
> The HSZ side looks to be the easy part, but I'm trying to
> put together a plan for deleting and replacing the relevant
> LSM volumes, disks, etc and updating the ASE configuration.
>
> What I'm looking for is what other people have done in
> similar tasks andif there were any gotchas, time savers
> (which is important), and like.
>
> Any information is much appreciated. I have my trusty manuals,
> but they don't always tell you everything!
Received on Wed Oct 13 1999 - 02:46:50 NZDT