Using Automation to Validate Applications and Services in Rubrik Backups
The ability to restore files, folders, applications, and entire servers is the ultimate goal of data protection. With that in mind, there are a lot of checks in place to validate that all source data has been correctly stored within the Rubrik Cloud Data Management (CDM) fabric. This includes CRCs to ward against corruption, over-the-wire encryption to prevent tampering, and the use of erasure coding to opportunistically place data across hardware components to avoid single points of failure.
In addition, it’s a good idea to validate that a backup can be restored. The best method is to perform a restore of the data to ensure that whatever is required – be it files, folders, applications, or entire servers – can be presented back to the users. Traditionally, this has been clunky. Solutions that offer such workflows are often quite slow to complete and limited in scope. Plus, they operate in a closed-source orchestration engine that is tightly coupled to the solution itself. Wouldn’t it be great if you could perform backup validation using your existing configuration management engines instead?
With Rubrik, validating is easy. Because the entire CDM fabric is designed to operate at scale across any on-premises or cloud environment, there is a powerful suite of RESTful APIs used by the HTML5 interface. Customers can access this suite at no cost and without licensing. You can very easily consume the APIs to add backup validation workflows to any existing orchestration engine with just a few steps.
As an example, I’ve written a workflow using PowerShell (one of my favorite object-oriented scripting languages) to plug into a testing framework named Pester. I’m already using this framework to test other parts of my infrastructure, such as the vSphere configuration with Vester, which keeps me from having to reinvent the wheel with yet another orchestration engine.
How does it work? Let’s start by examining the below screenshot, which shows the results of a Sandbox test on one of my SQL Server virtual machines.
Pester is being asked to run 10 unique tests. After all of the tests complete – including the creation of a Live Mount virtual machine, migration of the network, and application tests – the results show that all 10 pass. Because of this, one final API call is sent to Rubrik asking for the Live Mount to be discarded.
In 62 seconds, we’ve performed an end-to-end test for a backup without any human intervention. As a bonus, all of the details from the test are output as a JSON string that can be intercepted by a dashboard display or any back-end system designed to monitor, log, or alert based on data flow.
With Rubrik, you have the ability to easily integrate dynamic and modular tests to validate that your data protection investment is able to restore whatever you feel is important. And you can do this using your current (or future) orchestration engine, configuration management tools, and monitoring and alerting system. Imagine handing over that sort of capability to your application, operations, or automation teams and (without a doubt) know in a matter of seconds if the protected data is truly available! This is the power of treating RESTful APIs as a first-class citizen across an intelligent and scalable fabric.
Want to learn more about Rubrik’s APIs? We’ll be releasing an on-demand version of our webinar: Automatically Ensure your Applications, Services, and Servers can be Restored.