Security

Hi dears,
I have to upgrade a ClearPass Cluster, version 6.3.x to 6.5.0 version.

I think that I have to do the following:

1) Install the Patch 6.3.6 on all members (Need I to broke the cluster?)

2) Install on the publisher node the Cluster Upgrade tools
3) Download on the publsher node the 6.5.0 image.

I don't need to edit the virtual disk of the Virtual Appliance?

In the release note of 6.5.0 i found:
"VM only: If you have two disks already loaded with previous ClearPass versions—for example, 6.2 on SCSI
0:1 and 6.3 on SCSI 0:2—then drop the inactive disk before upgrading. You must then add a newer disk
based on the 6.5 disk requirements."

But in the Cluster Upgrade Tool Guide, there isn't any note about it.

Can you help to understand the correct steps to upgrade my cluster?

Thnaks in advance

Andrea

how big is just cluster? i have done a few small cluster upgrades and i don't use the cluster upgrade tool at all. just do the publisher first and then the subscribers, worked out fine.

as for the disks, if you started from 6.3 you are fine i believe. this is from earlier instalation when only one disk of 250 gb was added. if you have one of 500 gb you don't have to drop anything.

to add to boneyard - 

If your already on the 6.3 code train you have the new disk config layout so that issue is mute.

Yes, please ensure you upgrade ALL nodes to 6.3.6 [no need to break the cluster], and again as advised by boneyard, upgrade the PUB first followed by the SUB's in your cluster.

My recommendation is that you then run for 7-10 days on 6.3.6 to ensure everything is good before you plan your outage to upgrade to 6.5.0.

HI ,
the cluster is composed by two nodes .
I have upgraded it succesfully.

But after some days the appliance starts a particular behavior.

The publisher node goes down randomly and i have to reboot it becouse it became unreachable.

any ideas?

regards

Andrea

Based upon the symptoms you are better creating a TAC case.

Yes,

I opened a ticket some day ago.

I will update you with solution.

Regards

Andrea

did TAC come up with something Andrea?

Hi,
support doesn't found a root cause of the issue.

The issue seems to been solved aftet a shutdown of the Virtual Appliance.

We think that in this manner the database has been re-initialized.

Regards
Andrea