Hello,
I have to plan the upgrade a cluster of two HW appliances (C3000, HPE DL360 Gen 9) from 6.10 to 6.11
My customer concern is about the risks of a failing upgrade of an HW appliance as there is no rollback possibility (other than RMA).
If the upgrade to 6.11 fail on one of the servers he will be running the whole network with only one HW server !!
All Wifi & Wired devices (3000-3500 devices) are relying on Clearpass for network connectivity (802.1x & MAC auth).
I have seen that some peoples have purchase some spare disks to be installed on the appliances before the reimaging but
as the C3000 have 6 SAS disks it could be very expensive.
Also my customer doesn't want to migrate the production cluster to VM appliances as it would simplify the recovery.
My idea is to deploy, temporarily and only during the upgrade process, a third server on VMware to the existing 6.10 cluster
Like that if the upgrade procedure fail on the first HW appliance (Publisher) we still have 2 CPPM servers running (on 6.10) and we can
wait until a TAC case (RMA or fix) can restore the failing appliance.
If the first HW appliance upgrade goes well we have the same risk when upgrading the second HW appliance (Subscriber), then
we could deploy a new virtual CPPM Subscriber on 6.11 to keep a two servers cluster until the problem is resolved.
My concern is that I was thinking of using an Eval Platform license on the virtual server (that would be decommissioned after both
HW appliances have been successfully upgraded).
Is it possible to run a cluster with a mix of Permanent licenses (Platform & Access) and an Eval licence (Platform only) on the
VM appliance ?
I don't want to open a proactive TAC case for this topic right now since it will take a long time and also because HPE (as other vendors
like Cisco or Paloalto) are assigning penalties on Partners for non bug related cases.
Thanks for your advises.
Kind regards
Christian Chautems