NDMP incremental forever.

[heinkel]

Hi all.

My customer has a huge EMC Isilon NAS, containing 2.4 Pb of data. All is not backed up, but some folders are.

These contains typically up to 80 TB of data. A full backup will run NDMP backup to tape and take 2-3 weeks to complete, so we need to avoid full backups.

Users are filling folders when a project is finished, in order to save files to tape. RL = 5 years. The specific folder will then contain previously backed up data. So we will backup using a diff inc schedule. Users adds files and folders in a ad hoc way.

This works fine for a few days, after a FULL has run. And then it reverts to FULL and

We can see that backup level 9 has been reached, and a diff inc is previous backuplevel +1.

We want to run a diff incr forever, not having to use FULL schedules. How can we do that?

Best regards

[RiaanBadenhorst]

In NetBackup 8.0 there is a new stream handler for Isilon which basically means it will be able to benefit from aceleration. It should be released soon so I suggest you wait a bit, hopefully it helps with your issue.

[sdo · Accepted solution]

This doc suggests that NDMP on the Isilon can be instructed to use Level=10, i.e. incremental forever:

https://mydocuments.emc.com/pub/en-us/isilon/onefs/7.2.1/ifs-pub-backup-and-recovery-guide/07-ifs-br...

Try passing the Level=10 to the Isilon NDMP data mover by specifying at the start of backup selection directives in the backup policy.

[heinkel]

Thanks!

Implemented yesterday, and it seems to work. Will keep an eye on this!

[heinkel]

Interesting!

Thanks for that information!

[sdo]

The use of Level=10 is also mentioned on page 22 of:

    NetBackup for NDMP: NAS appliance information

...here:

    https://www.veritas.com/support/en_US/article.000027113