Create a Post
cancel
Showing results for 
Search instead for 
Did you mean: 
Robert_Sisk
Participant

Is there a way to do a hardware upgrade of an HA cluster without taking the site down (at all)?

Hi Everyone,

We have two model 5200 appliances in an HA configuration (R80.10 JHF 249) and we are planning to upgrade them to model 6600 plus appliances (running R80.40 latest GA JHF). Is it possible to upgrade the hardware without taking the networking down, say something like a mixed cluster upgrade?  SK article maybe or some other reference/procedure?

Any comments and/or experiences related to performing such a maneuver would be welcome!

Thanks,

Bob

0 Kudos
4 Replies
Robert_Sisk
Participant

We are planning to upgrade the 5200's to R80.40 before the swap, if that maters/helps.

0 Kudos
PhoneBoy
Admin
Admin

Downtime is inevitable if you're swapping hardware as there is no supported way to cluster different hardware types together.
However, R80.40 (and R81) supports multi-version clustering.
Refer to the Installation and Upgrade Guide for specifics.

0 Kudos
Robert_Sisk
Participant

Thanks for the reply. I was pointed to sk164212 as a possible option, which I believe you are referring too.

Multi-Version Cluster (MVC) replaces Connectivity Upgrade (CU) in R80.40 (sk164212)

In R80.40, Multi-Version Cluster (MVC) is introduced. MVC is a next generation upgrade method which replace the deprecated "Connectivity-Upgrade" (CU).

More information on MVC can be found in Check Point R80.40 Installation and Upgrade_Guide.

 

Check Point R80.40 Installation and Upgrade_Guide

Supported Versions in Multi-Version Cluster

The Multi-Version Cluster (MVC) in an R80.40 Cluster Member supports synchronization with peer Cluster Members that run one of these versions:

R80.30 (see sk144293).

R80.20 (see sk122485).

R80.10 (see sk111841).

R77.30 (see sk104859).

In an Multi-Version Cluster, the Cluster Members can run only these versions:

R80.40 and R80.30

R80.40 and R80.20

R80.40 and R80.10

0 Kudos
Scott_Bily
Participant

Curious what you ended up doing in the end?  I have a similar scenario but going from a cluster of 4800 to 6600s.     

0 Kudos