Skip to main content
commvault.com commvault.com
Question

Backups of VMs from one subclient may be subject to longer retention periods across different weekends

  • April 15, 2026
  • 3 replies
  • 15 views
M
Apprentice
Forum|alt.badge.img+5

Hi,

After migrating subclients from Index v1 to Index v2, VM backups within application clusters are losing time consistency on Selective Copies (e.g., First Full of the Month). The issue arises because Index v2 splits the subclient backup into individual child jobs for each VM; when these jobs finish after midnight, the system assigns them to different retention periods. Due to a fixed backup window, we cannot delay the start times, and manually selecting jobs for tape marks is not scalable.

How to configure it so that Selective Copy treats all child jobs of a given subclient as a single time unit, ensuring a consistent point-in-time recovery for the entire application cluster?

 

Kind regards,

3 replies

Sureshkumar S
Vaulter
Forum|alt.badge.img+12

Hi ​@mextrinsic ,

Good day!

This behavior is expected with Index v2. In Index v2, each VM backup is a separate child job, and Selective Copy jobs are selected based on the completion time of each child job, not the parent job. There is no documented configuration or setting that forces Selective Copy to treat all child jobs of a subclient as a single atomic unit for period selection.

Ref : https://documentation.commvault.com/2023e/commcell-console/jobs_and_job_control_for_vm_centric_operations.html

Regards,

Sureshkumar S

M
Apprentice
Forum|alt.badge.img+5
  • mextrinsicApprentice
  • Author
  • April 16, 2026

Hi ​@Sureshkumar S 

It is known that when using Index v.2, each VM is backed up as a separate job.

How can this be resolved so that backups for “cooperating VMs” are from the same time period? It’s quite important to keep backups from those same weekends.

 

Regards,

Sureshkumar S
Vaulter
Forum|alt.badge.img+12

Hi ​@mextrinsic ,

As per logic the retention should fall as per the backup start time.

We may need to look into your infrastructure to understand the scenario.

can you please raise a support case on this ?

Regards,

Sureshkumar S

Terms & ConditionsAccessibility statement