Rubrik : Check Backups Integrity

Background

Rubrik is good at doing backups, but in the enterprise world, backups are nothing if there are no restore tests planned. You know how boring this could be. Rubrik make backup great again by providing efficient and reliable sexy UI but the restore tests are still painful, time consuming and not appealing. Guess what ? Rubrik came again to another revolution : the backup integrity check !

Rubrik CDM 5.3 Backup Verification

Starting with CDM 5.3 (will be released somewhere in Dec-2020), Rubrik will propose a way to test the integrity of our backups. This is a mean to confirm that the data stored on the Rubrik appliance are matching the production data at the time of the backup. Based on checksums. It means there are no need to physically restore the data, just the need to compute the checksums and compare them.

As today, there is no UI for this yet, it will come later in future release. But you can easily use API calls to solved this modern world issue. Yes, I have a script for that ;)

The endpoint we are going to use is : 

https://<cluster_ip>/api/v1/backup/verify

In a nutshell, this is another async process : first ask Rubrik to verify a specific snapshot, and later check the result. Very easy to use.

Remark : this process is CPU intensive and even running in the background, this is not recommended to run it when other tasks already running. Ideally, run it outside of your backup window if any. You will feel the CPU load on a EDGE raising sky, the behavior on a physical appliance if of course different, but still use it with caution.

Before starting using this feature, you need to enable it. Two options here, either you have rksupport access to your Rubrik cluster, either you need to call Rubrik support.

If you have access to the cluster, just issue this command to enable the flag :

~$ rubrik_tool.py update_config blobstore isBackupVerificationEnabled true

~$   

~$rubrik_tool.py get_config blobstore isBackupVerificationEnabled

true

When done, you can access the API endpoint related to backup verification. So far, I could verify vmware VMs, Nutanix AHV VMs and FileSets. At the time I'm writing these lines, there are the only object (in addition to the managed volumes) that we can verify.

The code is part of my Rubrik Framework and is available on GitHub. The two new functions are :

function rkIntegrityResult($clusterConnect,$eventID)

function rkStartIntegrityChk($clusterConnect,$objectID,$snapID="")

rkStartIntegrityChk initiate the verification process. This one is a one shot execution.

rkIntegrityResult get the status of the request. This one will be looped to verify the progress.

There is a code snippet as well on GitHub showing how it works.

Here is the sample output of a Nutanix VM verification (speed up, real time is 6 mins) : 

I hope this will help others to meet their ISO requirements ;)

Oh, and a big thanks to Jaap Brasser for his support !