Switching a specific VS.

[Matlu]

Hello,
I have an MDS environment with VSX.
I have VSX clusters in VSLS mode, and I want to know if it is possible to switch only one specific VS.
For example, in a VSX cluster, I have approximately 10 VSs that have their VS standbys.
My VSX Node 1 is the one that currently has all VSs in ACTIVE mode, but I only want to “switch” the order of VS ID 3. Is this possible?
I want all other VSs to remain active on Node 1, but I only want to switch 1 of the VSs.
I have a CMA that manages this VSX box.
Thanks for your comments.

[the_rock]

vsenv?

[JozkoMrkvicka]

Since you mentioned you are using VSX with VSLS mode, you can perform failover on each VS, without impacting clusterXL state of other VSs.

Switch content on VSX Node 1 into VS ID 3 and perform failover. Failover will be done only on VS ID 3.

Before failover, make sure both members of VS ID 3 are in ACTIVE - STANDBY state.

To perform switch (failover) to only on VS ID 3 using clish (perform commands on VSX member where VS ID 3 is ACTIVE):

> set virtual-system 3

> set cluster member admin down

> set cluster member admin up

To perform switch (failover) to only on VS ID 3 using expert mode (perform commands on VSX member where VS ID 3 is ACTIVE):

# vsenv 3

# clusterXL_admin down

# clusterXL_admin up

[Matlu]

Hello. @JozkoMrkvicka 
This will only affect VS ID 3, correct?
Will the rest of the VSs retain their position, remaining “ACTIVE” on Node 1 and “STANDBY” on Node 2?
Thank you.

[JozkoMrkvicka]

As far as you are really using VSLS mode of VSX, only VS ID 3 will be affected.

[Matlu]

Yes, I am using this VSX Cluster mode.

[Expert@FWCP-01:0]# cphaprob state

Cluster Mode: Virtual System Load Sharing (Primary Up)

ID Unique Address Assigned Load State Name

1 (local) 10.0.0.25 100% ACTIVE FWCP-01
2 10.0.0.26 0% STANDBY FWCP-02

Active PNOTEs: None

Last member state change event:
Event Code: CLUS-114904
State change: ACTIVE(!) -> ACTIVE
Reason for state change: Reason for ACTIVE! alert has been resolved
Event time: Fri Sep 12 14:02:29 2025

Last cluster failover event:
Transition to new ACTIVE: Member 2 -> Member 1
Reason: Interface Mgmt is down (disconnected / link down)
Event time: Fri Sep 12 12:37:52 2025

Cluster failover count:
Failover counter: 1
Time of counter reset: Fri Sep 12 12:37:05 2025 (reboot)

Cluster name: FWCP-00

Virtual Devices Status on each Cluster Member
=============================================

ID | Weight| FWCP-01| FWCP-02
| |
| | [local] |
-------+-------+-----------+-----------
3 | 10 | ACTIVE | STANDBY
4 | 10 | ACTIVE | STANDBY
5 | 10 | ACTIVE | STANDBY

Can I be certain that it will not affect other VSs according to my scenario?

[genisis__]

You should also be able to do this via vsx_util vsls command on the CLI of the MDS.  

[Chris_Atkinson]

@Matlu is this a temporary change or do you want VS3 to prefer the other node on an ongoing basis?

https://sc1.checkpoint.com/documents/R82/WebAdminGuides/EN/CP_R82_VSX_AdminGuide/Content/Topics-VSXG...

[Matlu]

Hello,

This is a temporary change.

We just want to switch the original order of VS ID 3.
Cheers 🙂

[the_rock]

For what is worth, here is the AI response to it.

Andy

*****************

ChatGPT said:

Yes — in a VSX cluster running in VSLS (Virtual System Load Sharing) mode, you can fail over (or “switch”) a single Virtual System (VS) to the standby member, without affecting the others.

Here’s how it works and what you can do:

---

✅ Behavior in VSLS

• Each VS has its own Active/Standby state per cluster member.

• Failover events are per VS, not necessarily cluster-wide (unlike traditional HA mode).

• You can move just one VS to the other node while keeping the rest as-is.

---

🔹 How to switch one specific VS

You can use the ClusterXL commands with the -v option to target a specific VS:

1. Log into the VS context of the cluster member that currently owns the VS you want to move:

   set virtual-system 3

2. Issue the failover command:

   cphaprob fo

   → This forces the VS to fail over to its standby peer.

   Alternatively, you can use the clusterXL_admin command directly on the VSX gateway in the context of that VS:

   clusterXL_admin down -p <vsid>

   Example for VS ID 3:

   clusterXL_admin down -p 3

   This will move only VSID 3 to the other cluster member.

3. Verify status:

   cphaprob stat

---

🔹 Notes / Considerations

• Other VSs will remain untouched and continue running on the same node.

• If you use cphaprob fo without the -v or without entering a VS context, you risk failing over the entire cluster or all VSs.

• Make sure that both members are fully synchronized before switching (check cphaprob stat or vsx stat).

• This operation is commonly used for maintenance, balancing load, or troubleshooting.

---

✅ Answer: Yes, it’s possible. In VSLS mode, you can fail over just one VS (e.g., VSID 3) to the standby cluster member while keeping the rest on the current active node, using clusterXL_admin down -p 3 or by entering the VS context and running cphaprob fo.