Rubrik : RPO in Minutes Instead of Hours

Background

When you have very critical applications, sometimes, the minimum of 1h RPO (Recovery Point Objective) that can be configured on a SLA is not sufficient. There is a way to specify a SLA with minutes instead of hours. But, for this you need API calls !

Let's API

Indeed, I had the need to take a snapshots every 30 minutes for some specific application requirements. Then I've found a way to do it with API calls.

Warning : Indeed, if Rubrik is not able to take the SLA at the requested interval, you may fall into some bootstrapping problems and a never ending jobs. I recommend to do some tests before implementing this into production.

Simple way to create a SLA that occurs every 30 minutes and keep the snapshot for 3 days (72 hours) :


rkCreateSLA($clusterConnect,"API_Created_SLA",30,72,1,1,1,1,1,1);


In order to run this command, you need to deploy my rkFramework.php found here.

This will create a SLA visible in the Rubrik UI but the UI is a bit broken



The reason why the values seem wrong is because the UI is not ready yet to take this into consideration but the API stack is. In short, there is a good reason to think that Rubrik is testing intensively this feature but is not ready to release it yet. As I'm writing these lines, current Rubrik CDM is 5.0.1-p1-1342. We can have good hope that version 5.1 will support minutes-based SLAs.

Checking what's in the Rubrik configuration using API call confirms that we have a valid SLA defined with the value we wanted :


It means that you can still try it using API calls, it works fine I have tested the above SLA on my test platform. But if you want to have full visibility on the UI, you'd better wait a little bit.

I hope this helps !

Comments

  1. How is this easy or sexy!? Looks like old TSM commands to me. Go Veeam, can do all the same things at 50% the cost.

    ReplyDelete
    Replies
    1. You have the right to say that. I'm not sure this is relevant but you still have the right. Veeam is not coming with storage, so comparing cost between solutions never been an easy exercise.

      Delete

Post a Comment

Thank you for your message, it has been sent to the moderator for review...

What's hot ?

ShredOS : HDD degaussing with style

Nutanix : CVM stuck into Phoenix