topic Re: VSX R80.30 (VSLS cluster) with 4 members question in Firewall and Security Management

VSX R80.30 (VSLS cluster) with 4 members question

KostasGR — Thu, 18 Nov 2021 11:37:34 GMT

Hello community

On my lab I have a cluster VSX R80.30 (VSLS cluster) with 4 members.

Also i have configured a Virtual System that is Active on VSX A , Standby on VSX B and Backup on VSX C and Backup on VSX D.

What is the recommended way to stop the VS from being even backup on VSXC and on VSXD and instead being on DOWN state.

From skes i have read that the recommended way to failover is through vsx_util vsls but the above is something different.

The other way i have thought is to login on Virtual System on VSXC and on VSXD and execute clusterXL_admin down.

After that when i want the VS to be in BACKUP state again execute clusterXL_admin up on VS on VSXC and VSXD.

Do you see any impact on the above steps?

BR,

Kostas

Bob_Zimmerman — Thu, 18 Nov 2021 14:00:41 GMT

At a fundamental level, the impact would be that some VSs can only handle traffic on some members, but that seems like what you're trying to do.

What exactly is your goal? I would consider a cluster with an active/standby/down/down VS to be broken.

KostasGR — Thu, 18 Nov 2021 14:29:28 GMT

Consider a scenario that VSX sync interface between Site A (VSXA VSXB) and Site B (VSXC and VSXD) is lost on layer 2.

So i want the active /standby VS on site A not been impacted at all.

BR,
Kostas

Bob_Zimmerman — Thu, 18 Nov 2021 14:45:17 GMT

At that point, why have VSXC and VSXD in the same cluster? Just make two clusters and run the same policies on both.

genisis__ — Thu, 18 Nov 2021 20:47:29 GMT

I agree.

Not really used this but weights could be explored, but in all cases I would not purposely leave a VS in a down state on a VSX appliance.