Most people will look at the above graphic and focus on those big, beautiful Andes mountains – and rightly so – The massive snow-covered peaks are pretty awesome – just like some of the big, massive new features getting coverage from the Rubrik Andes 5.0 release. But I say, let those mountains represent things like Oracle Instant Recovery, SAP Hana support, NAS Direct Archive, and Elastic App Service. Let those big new features like EPIC EHR Support, Windows Bare Metal Recovery, and Office 365 Protection steal the show. But what about those shining stars in the background? What role do they play? Aside from twinkling and being super awesome, I’d say they represent a ton of the smaller, awesome, less-covered features that get bundled in with a major Rubrik release. And in my opinion, it’s these smaller features, these subtle yet super useful improvements that really drive a release. With that, let’s have a look at some of these real stars of the show <- see what I did there 🙂
This is the number one feature in 5.0 in my honest opinion – sure broader coverage is cool, expanding what we can protect and all that – but the SLA Domains, the data protection policies, live at the core of the Rubrik product! Applying policy, rather than creating jobs is a real driver in the simplicity of using Rubrik. It allows us to move away from the imperative approach of creating specific jobs and move into a more declarative way of defining end state – I want to back up this, every x hours, every x days, every x years, archived here, replicated there, etc – no matter what the workload, no matter what the application. This is where Rubriks simplicity shines! But there is a fine line between providing simplicity and accommodating the multitude of scenarios that pop up in any given environment. That said the new SLA enhancements have done just that – granting backup administrators further control over when their data is backed up while still maintaining the core design of being simple and easy to use. Andes brings choice when it comes to SLA’s. The choice to select whether you want monthly backups to occur at the beginning, middle, or end of the month. The choice to process yearly backups on either the first or last day of the year – and better yet, the choice to state when your fiscal year begins. Options to create new weekly backups occurring on whichever day you specify, and even quarterly backups as well. This is a big enhancement to the product for those looking to have a little more control over when backups occur. On the flip side, for those that take advantage of Rubrik’s smart task scheduler and let CDM do the hard work of figuring out when to back things up, there is absolutely no change – customers can choose to take advantage of the new SLA enhancements, or simply leave them as is.
Rubrik now offers more control over what happens to backups after objects are no longer being protected. What does this mean? Well, you can now apply SLA Retention policies to unmanaged objects – allowing customers to state how long they would like to hold onto backups for things like deleted VMs and databases. On top of this, customers can now instruct Rubrik to simply archive these backups off to the public cloud or local NFS, freeing up valuable space on the Brik for other backups. And of course, if you wish to simply just remove any old backups from an unmanaged object, the ability to simply delete all the backups is still available. The point is, customers have full control over how long their data is stored and where it is stored, whether the data still exists in their production environment or not.
SQL Change Block Tracking
In the case of VMware, a native Change Block Tracking (CBT) mechanism is provided which tracks and reports on which blocks have changed within a given VM since a previous point in time – This allows Rubrik to simply tap into CBT and pull just the blocks needed in order to create an incremental backup. Unfortunately there is no such native functionality within SQL Server – meaning databases need to be manually processed in order to determine changes. That’s not to stop Rubrik though. The 5.0 release brings with it a custom-built filter driver that can be deployed, allowing Rubrik to track changes as they are made to the databases residing on the SQL Server. This CBT filter driver makes backups a breeze as Rubrik can simply poll the filter driver for the changed data and process it accordingly – in the end, we are left with the ability to lower our RPOs and take more frequent backups of our databases! Snazzy!
SQL Server Group Backups
In the event that you have a number of databases residing on the same SQL Server, all which belong to the same SLA Domain, Rubrik will now group these databases together when it comes time to take a VSS read-consistent snapshot for those databases – what does this mean? Well, instead of taking a snapshot per database (imagine the time that may take if processing hundreds of databases), now only one snapshot is required! Again, a focus on lowering customer RPOs as well as overhead on the production SQL Servers.
Per VMDK VMware Live Mounts
Because you can never have too many options at your fingertips when you are in a scenario when you are in a recovery situation. VMware VM recovery on a “entire VM” basis has been around since the day Rubrik came out of stealth, however, 5.0 brings the ability to drill further down into this and Live Mount a single VMDK from VM protected by Rubrik. This means we can take a single virtual disk from a VM and live mount this back to the original VM, or any other VM of our choice – all the while not affecting the immutability of the backup itself. But what does this really mean to you? IMO – options, and a faster recovery time! Say you have a 10TB VM containing 2 virtual disks (One is 9.5 TB and the other 500GB). If you experience corruption on the 500GB drive there is no need to recover the entire 10TB just to gain access to the 500GB you care about – simply live mount the 500 GB drive! Nifty!
This is not necessarily a new feature, but a classic example of “we provide way more in the API then we do in the UI”. We’ve long been able to recovery vCenter through the API by exporting it to standalone host – but 5.0 brings this feature front and center by adding it into the UI. The problem addressed here is a classic chicken and egg scenario. If you virtualize your vCenter, how do you recover your vCenter when your recovery target points at a vCenter – wow, that’s a lot to take in. You will never get into that virtualized vCenter, what comes first, chick or egg scenario with Rubrik as you are now able to set a standalone ESXi host as a recovery target during the recovery and export options!
API Token Generation
Being an automation junkie this feature is near and dear to my heart. I write a lot of scripts, scripts which I love to share with others – the problem is I’m constantly having to comb through those scripts removing any hard-coded usernames and passwords before they get shared – or worse, building complex functionality around how to input the username/password and having to explain that to anyone utilizing the script. The new API Token Generation functionality in Rubrik seeks to remove those exact problems. Rather than hardcode passwords in scripts one can simply use a variable representing an API token. The token itself is generated from with the Rubrik UI, granted specific permissions, with the ability to add an expiry date to revoke access. This is a great way to promote the sharing of scripts and knowledge built around Rubrik – and not only that, provides administrators a means to easily revoke access to any script that might be running against their Rubrik cluster by simply removing the token.
These are just a few of my favorite outlying stars surrounding the Rubrik Andes 5.0 release. There is certainly much much more in regards to LDAP support, Syslog enhancements, banner customizations – this article could go on forever! If you want to learn more about these features, or anything else Rubrik related feel free to reach out – that said, also keep your eyes peeled on Rubrik.com and sign up for this webinar series which is going to dive into some of the Rubrik Andes 5.0 functionality a little deeper.