ClusterXL Improved Stability Hotfix

[Danny]

For everyone who is experiencing Cluster instabilities during policy installation, caused by interface flapping events or high CPU load Check Point released a ClusterXL Improved Stability Hotfix for R77.30 & R80.10 (sk133372).

The fix is not integrated yet within any Jumbo Hotfix GA-Release and needs to be installed separately.

Code improvements:

• Cluster policy skip in case no change was done in cluster configuration.
• Freeze mechanism during cluster policy installation has been improved.
• New grace mechanism to reduce risk for failovers due to flapping interface or sporadic CCP loss.
• Failover is not triggered due to SYNC interface disconnection.
• Link down is now detected quicker to achieve better failover response.

[Kaspars_Zibarts]

Aha, interesting as we had sudden spit of ClusterXL issues recently on some gateways after happily running take 42 for 9 months. Upgraded to take 112 in July and it seemed to settle it (take had few fixes for ClusterXL). But it seems like there are more gremlins out there. Thanks for sharing!

[JozkoMrkvicka]

Is this stability hotfix integrated into R80.20 GA release ?

[Jeff_Engel]

Yes, this hotfix is integrated in R80.20.

[Corey_Christmas]

Anyone had luck with this hotfix on R77.30? We just experienced an Active/Active split brain outage on an HA pair after a policy install. We are already on HFA 338.

[Alexander_Wilke]

Hi,

we are on JHFA Take 336 and we still get notifications that partner member is going down or interfaces are down when doing the policy installation. My hope was that this will be fixed with this Hotfix but I don't think it is.

Regards

[Corey_Christmas]

Are you going to try the hotfix anytime soon?

[Dawei_Ye]

It seems Check Point didn't release this hotfix for R77.30 nor ingrated it into the latest jumbo hotfix.

We are facing the same scenario without solutions.