Skip to main content
Question

missing index backups

  • September 18, 2025
  • 4 replies
  • 42 views

downhill
Byte
Forum|alt.badge.img+10

hey guys,

All clients are v2 indexing at this point. I get an alert every so often that index backups haven’t ran in 30d for NDMP client. Each and every day there’s 100GB of inc. data being backed up, but the system scheduled Big Data Apps Index backups for that SP show that only “full” backups ever pick anything up, each day of the week when the incs run for the subclients of that NAS client, there doesn’t seem to be anything backed up at the index server level. Of course you can’t see any of this in the Command Center but is it normal for the classicIndexInstance backups to be zero bytes (and thus immediately age) in between Fulls?

I see this phenomenon even on very small, minimal churn non-NDMP clients, so my question is - is there some threshold which determines IF an index server backup has anything new to write? For my very large-ish VMware SP’s and their Big Data Apps Index backup - those scoop up stuff every day, there’s never a 0 byte index backup, then again the SP has 3300 VMs associated. So there must be some algorithm or minimum delta index change which determines whether it does anything or not??

I don’t see anything in the BoL for this nor anyone else seemingly asking this (or the search failed me).

thanks

4 replies

Forum|alt.badge.img+10

Hi ​@downhill ,

 

Good day!

  • If the index is not in a healthy state, it will not be eligible for backups under BigData Apps.

  • If you encounter this error for a specific database, try browsing or restoring the subclient/VM to verify the index health.

  • IndexServer clients (BigData Apps clients) are created automatically after the first V2 backup runs in a storage policy.

  • It’s acceptable to have a zero-byte backup in BigData Apps for a single day, but not consecutively if V2 clients are backed up under that storage policy.

  • Ideally, an index becomes eligible for backup once every 3 days or after 100,000 object changes, whichever occurs first.

Regards,

Sureshkumar S


downhill
Byte
Forum|alt.badge.img+10
  • Author
  • Byte
  • September 19, 2025

Hi ​@Sureshkumar S there are no errors or events logged that I can see. So if it goes a week without any such backup, do you think I should open a ticket to confirm? To me the 100k objects would absolutely not be met for this SP and the other I found with similar skipped backups. 

thanks


Forum|alt.badge.img+10

Hi ​@downhill ,

Yes, you can raise a case on that event to validate.

Regards,

Sureshkumar S


downhill
Byte
Forum|alt.badge.img+10
  • Author
  • Byte
  • October 13, 2025

I did open a case. It was determined that the source of truth (Index health report in Command Center) is telling us it’s OK. But, the alert still keeps nagging me with the same thing - no backup in last 30d. To me it smells a bit weird.