Central Deployment Upgrades a Cluster - Full Connectivity Upgrade

[Michael_Horne]

Hello All,

I am investigating the installation of hotfixes via the Smart Console and have come across one confusing piece of information. I assume there is just a unintentional use of words.  When I looked at how the central upgrade is done for a cluster it mentions that is used a "Full Connectivity Upgrade":

Operationally I have been following the "Zero Downtime upgrade" process for upgrading a Security Gateway Cluster. I wanted to understand the difference between :Full Connectivity" and "Zero Downtime".  Found something in the SK107042 that confused me: ClusterXL upgrade methods and paths (checkpoint.com)

Full Connectivity Upgrade seems to be obsolete since a very long time.

For deploying hotfixes and upgrades centrally does "Full Connectivity Upgrade" = "Zero Downtime"?

Many thanks,

Michael

[the_rock]

Personally, never tried this method...I always stick with zero downtime upgrade method, as it never failed for me.

[Scott_Paisley]

We have our clusters set for 'Switch to higher priority cluster member' upon cluster member recovery.

In R81, when you do a version upgrade from smartconsole, it upgrades the standby member, fails over traffic, upgrades the other member, and fails back.

In R81.10 this is no longer a permitted configuration. The upgrade from smartconsole will fail unless you turn off the cluster setting.

[the_rock]

I had seen so many issues with that setting, not only with CP, but other vendors as well, so I avoid it.

[Scott_Paisley]

That is the default setting across our entire estate. There was a good reason for doing it 5 years ago, but maybe we need to revisit it now.

[the_rock]

I understand. Personally, I find it used to work better in older versions though.