Rubrik: At it again with their Firefly release.
If you’re running a datacenter today and are at all involved with the storage side of the shop, chances are that you’ve heard about Rubrik, a backup company that really started to pick up steam last year. This year, they’ve picked up right where they left off… When you step back and look at the product two things hit you. First, it’s designed to make backup, something that’s traditionally been a total pain in the tail, easy. Secondly, it’s designed around scaling out, much like all of the big internet players today – think Google/Facebook. Scaling out, means that from a business model perspective you’re more capable of keeping your costs in line with your actual need/spend in your datacenter. Amazon likes to make a big deal about this and has this VERY famous graphic around demand and resources:
Think of it like this, because you’re always just buying a set node type and config to scale, your costs are closer to that demand line. Because you’re not spending a ton of time on configurations and sizing when purchasing Rubrik nodes you acquisition cycle is much quicker. Lower and easier acquisition time plus scale without lift and shift migrations between software and storage platforms are an automatic win in an industry mired by complexity.
Company Health – Financial Outlook
With any startup company generating cash to enable the company to execute on product development and strategy is key. Companies look for seed funding and then funding through other rounds to develop product features and, for paying customers, to keep the lights on.
Rubrik, over a very short history, has been led by strong investment by very prominent venture capital firms Lightspeed Venture Partners and Greylock Partners. In round “C” of investment, Rubrik is happy to bring on Vinod Khosla (Co-founder and CEO of Sun Microsystems) who’s a part of 61 million dollar round of investment in the company. Perhaps more notably here, Rubrik is now a repeat customer (Greylock/Lightspeed) with other firms who have already invested. While not totally unheard of, this pattern of re-investment by initial firms should be viewed favorably as it shows initial investors confidence in not only the product but the business model and business leadership.
For me, the latter is probably something that I would consider the best indicator of company viability. If multiple VC’s are re-investing, it’s easier to have more confidence in the overall business viability of a company. Looking at Rubrik, the general view is that the company appears to be quite healthy from a “do we have cash to operate” perspective.
Company Health – Growth Outlook
When looking at new companies and new tech it’s always amazing at how quickly great companies grab the market and run with it. When I think of Rubrik, the thing that really stands out to me is the overall pace of this growth. I’m a big running fan and am always amazed at the times really really great 400 meter runners can run (one single trip around a standard track)… When I’m training and doing speed work for longer races it always feels like my hair is on fire (pun intended) when I run anything quicker than a 75 second 400. Olympians can ease into that race by running a 46 or 47 second race (a whole 20 seconds faster than me… good gosh). Rubrik, like Wade Van Niekerk, is running the 400 of their lives… Since their founding in Q1 of 2014, Rubrik has moved through four major releases of the platform which are only getting better today. In their third year as a company, they’ve been named Gartner cool vendor, raised more that 150 million dollars, and grown from 4 to over 160 Employees. In general Rubrik is looking less like backup space competitor and more like Michael Johnson wearing gold shoes…
Archival Options – Now with Azure!
NFS, Amazon S3 or any S3 compatible object store have been the standards for Rubrik’s archival target up until Firefly. Now added is support for archival to cheap and deep Microsoft Azure Blob storage. Looking to spend those free Azure credits that have been thrown your way in a past Microsoft ELA? This might be a nice way to do it for your organization. Here’s a look at how complicated it is to configure Rubrik for Azure Blob storage archival target:
In my mind, the story here is less about the capabilities of being able to support both Amazon and Microsoft clouds and more about being able to support multiple vendor platforms quickly. In fact, the best part of the platform might not be the Azure or S3 options, but instead it might be the Object Store option, which really gives you access to leverage ANY S3 compatible service. Beyond that, if you’re truly looking to have support for the platform, I’d recommend actually talking to Rubrik, as they have shown that they can quickly certify and support object stores… Here’s a nice eye-chart of the compatible platforms that they are supporting today beyond Amazon S3 and Azure Blob:
EDGE – Real backup for your remote locations!
Edge is a great new option that is rolling out as a part of the Firefly release and frankly one that’s perhaps most interesting to me. Edge gives you all of the great software backup capabilities of Rubrik in a VM. The goal of EDGE is to provide remote offices a backup VM appliance that can ship data back to a Rubrik mothership cluster back in the datacenter. Edge provides the features of the Rubrik platform in a small package making this a great option for ROBO sites. The goal of EDGE is to give customers a local entry-point into the backup system so that backup data is managed to SLA and shipped in the most efficient way possible over your corporate network. A small caveat to using EDGE: while the whole use a VM appliance thing is nice, EDGE should never be deployed on the same storage that it’s backing up. Rubrik’s filesystem on the physical platform makes this a non-issue, but it’s an important global design decision at remote sites to ensure that this appliance truly has air-gap separation from the storage that it’s backing up.
Erasure Coding – How about double the space?
I like free stuff. You like free stuff. We all like free stuff. So Rubrik went “Oprah’s favorite things” on their customer base by adding support for erasure coding the platform. Erasure coding can be pretty technical when you get down to the nitty-gritty, but it’s safe to say that under the covers your data blocks are being spread out much like a disk array would do in a RAID-5 setup (using some calculations to rebuild the actual blocks of data if needed in the event of a failure). This is not a new technology in the marketplace, and as the theory has been proven at scale with other vendors, Rubrik has decided to enable the feature in its filesystem. Net-net: existing customers get the benefit of the additional space that leveraging erasure coding vs mirroring data between nodes allows for. BAM, nearly 2x the local capacity for FREE!
Physical SQL and Physical Linux
Rubrik is driven by an intuitive SLA policy engine. Basically, you tell Rubrik at what interval you’d like to be able to recover and Rubrik manages the backups to that SLA in the background. I’m very excited to see this simplicity extended to physical Microsoft SQL Server and Physical Linux. On the SQL side, Rubrik will allow for database level backups and more importantly, will manage your DB logs to that SLA for you. Selecting an SLA for a database goes from quite complicated to answering a couple of questions: How often should I back the DB logs up for you? How long should I keep your DB logs? Because this is extended to a physical host, a process (read: agent) does need to run on your physical server. This process manages working with the database and underlying operating system to ensure a clean and consistent backup and like all other features of the Rubrik platform is driven via API as needed by the Rubrik appliance.
Rubrik is taking this same physical approach and extending it to the Linux platform. File level recovery for physical linux hosts is now possible and aligns with your intended SLAs. This in my mind is simply step one to being able to support more and more types of recovery for each of these platforms and an area that we’ll need to keep watching in future releases.
Super Automatable – that’s a word right?
No need to go into how awesome the support Rubrik has for nearly all of the relevant enterprise automation/orchestration tools out there. PowerShell, Chef, Puppet, Salt, vRO, Ansible, there are people who are working with this platform who have developed and shared what they’ve used for nearly all of them. The important takeaway here is that because the whole platform was developed from the API up, creating a bit of code to allow a third-party tool to manage your SLA’s and adding/removing objects from backup management is easy and will continue to be. Being simple and extensible are core tenants of the Rubrik platform itself, it’s natural that these properties extend to the programmatic interfaces.
Keep in touch!
This is simply a general recap of some of the big features that I see in Rubrik’s latest release prior to the big show at VMworld 2016. I’d encourage you to keep an eye out here as I’ll be testing some of these new features in my homelab and provide clarity to the areas where I have admittedly been VERY general in my analysis. Until then, make sure you check your backups nightly for failure, or you could just grab some Rubrik, and not worry about it as much.