Skip to main content
commvault.com commvault.com
Question

How the CS hostname mechanism works during the failover process.

  • May 14, 2026
  • 1 reply
  • 19 views
K
Challenger

Hello Commvault Community, 

I’d like to ask how the CS hostname mechanism works during the production failover process. 

 

  1. If both CommServe servers, passive and active, have different “CommServe Hostnames” and the CommServe Hostname value isn’t changed for Client Computers (production servers) after a production failover, is this the correct behavior, or should this parameter be automatically updated in Client Computer Properties?

    CS Active: CS_ABC (Instace001)
    FailoverNode1: Fo_ABC (Instace002)

    CS Passive: CS_XYZ (Instance001)
    FailoverNode2: Fo_XYZ (Instance002)

    After a successful production failover to CS_XYZ, the environment works correctly without any problems, but all Client Computers still have CS_ABC listed as the CS Hostname in their properties (it acts as passive CS after the production failover). Is this how it’s should work, or should the CS hostname be automatically updated on all Client Computers as part of this production failover process? 
     

  2. I haven’t found any such description in the documentation, but I suspect that the communication and configuration logic of CS LiveSync are handled by the failover nodes (Instances 002 on both CS servers). What if one of the failover nodes (Instance002) becomes unavailable and the Commvault config points to the old CS hostname, which currently acts as the passive CS? Is this intended behavior of the configuration correct, or not?

    For example: (after successfull production failover process)

    - “CS_ABC” is active CommServe;

    - All clients have the CS hostname indicated in properties: “CS_XYZ”

    - Instance001 of CS_XYZ is offline (it’s normaln - CS LiveSync config)
      Instance002 of “CS_XYZ” is running (Up), so Client Computers can reach “CS_ABC” via Failover node.

     

    But what if CS_XYZ completely fails (instance002)? Then all Client Computers will try to reach a node that is not responding?

 

Thanks in advance!

 

Kind Regards,
Kamil

 

1 reply

L
Vaulter
Forum|alt.badge.img+19
  • LedoespVaulter
  • Vaulter
  • May 14, 2026

Hello Kamil,

Point1.

https://documentation.commvault.com/11.40/commcell-console/configuring_multiple_commserve_hostnames_for_disaster_recovery_01.html

Clients know the names of the CommServe prod and standby and they know who is in the one active due to sCSInterfaceList setting. Important the clients are in 11.14 or higher to have the intelligence in the software to understand this. No need to change the CS hostname on the clients.

 

Point 2.

When you enable CS Live Sync you can specify to use as network proxies the Failover nodes, this is optional. If enabled then the second instances are the intermediates or proxies that talk with the active CommServe (both talk and are reached by clients). If not enabled, then clients talk with the active CommServe directly. 

If Failover nodes are configured as proxies, clients will talk with CS_ABC and CS_XYZ (second instance), if one of them is down then will talk only with the remaining one. So no matter what is the CS hostname defined for the client, it does not mean they will talk with the second instance named with same CS hostname, they talk with all Failover nodes, in your example two.

Terms & ConditionsAccessibility statement