My readers know that I have a weak spot for Pure Storage as well as Cohesity, as I believe both companies are innovators and disruptors of their own markets. I had the occasion to be briefed by Cohesity on their most recent platform improvements in May 2017 at their HQ, then I had also the opportunity to learn news from Pure Storage while participating to Storage Field Day Exclusive at Pure Storage Accelerate, even if those news did not relate to Cohesity FlashProtect.
While this wasn’t covered during Pure Accelerate, one of the most interesting features I’ve seen was the integration between both companies’ products. Let’s look at it, but before we start, let’s have a quick introduction for readers who might not know about Pure Storage and Cohesity.
Introducing Pure Storage and Cohesity
If you haven’t heard about Cohesity before, I recommend you read this article I wrote last winter. Cohesity is a solid startup that has built a broad secondary storage platform. It encompasses many use cases, and while it certainly qualifies as a data protection solution (either as a backup target, or with the addition of their own data protection solution) there is much more than can be achieved, including Data Analytics. Customers can begin with one feature, then start using additional functions of Cohesity’s Secondary Storage platform as they need, allowing them to adopt the technology non-disruptively, at their own pace.
Pure Storage, if you don’t know them, are an innovative publicly-traded storage company that has led the adoption and democratisation of all-flash arrays in the data center world. Their focus is on all-flash primary storage for mission-critical workloads as well as high-performance object storage for applications that have high-throughput and low-latency requirements. If you want to learn more about them I recommend you read my articles on their FlashArray //X NVMe all-flash array platform as well as on Flashblade, their all-flash object storage solution. I’ve also written a bit about real-time analytics and machine learning in Pure Storage’s context.
Cohesity FlashProtect: how does it works?
Pure Storage arrays already have a built-in snapshot feature which allows for fast RPOs and RTOs without impacting performance. This snapshot feature, while efficient, doesn’t offers tiering in terms of data retention policies and snapshots are only present on the array serving primary storage requirements.
The addition of Cohesity as a secondary storage and data protection solution allows for a deep integration with Pure Storage snapshotting capabilities. This integration is done by leveraging Pure Storage APIs and manifests itself by offering customers with a single interface that allows to manage multiple tiers of snapshots, whether on the primary storage (the Pure Storage FlashArray device, for example), but also on the Cohesity platform and eventually on any other tier, such as cloud storage if needed. An obvious advantage is therefore the ability to restore Pure Storage consistent snapshots to another Pure Storage array (located in a DR data center or a secondary site) in the unlikely case of a complete data loss (hazard/disaster in the primary data center, for example).
The lifecycle of snapshots is managed through data retention policies created in the Cohesity platform. Those policies control the creation, retention and eventual deletion of snapshots in an automated way at every layer, i.e. on the Pure Storage array, on the Cohesity appliance, and in the cloud. Because Cohesity supports Amazon AWS, Microsoft Azure and Google Cloud, and if I’m not mistaken also any S3-compatible cloud provider, customers are given enough choice and flexibility in their cloud long term retention options. Finally, Cohesity have added the ability to execute pre/post snapshot scripts to ensure data is quiesced before the snapshots are taken, with the goal to obtain crash-consistent snapshots where the integrity of the data is guaranteed.
This collaboration between Cohesity and Pure Storage is not only logical, but also important for both companies. First of all, both companies are in some sort the « underdogs » of their markets (even if I’m extremely over-simplifying): they’ve undoubtedly made their proof from a technical and engineering perspective, but they still have the world to conquer.
Pure Storage has been consistently gaining momentum but still has a lot of opportunities/market shares available to grasp considering the heavyweights of storage that are DELL EMC, HPE and NetApp to cite just a few. Cohesity is leading the battle on two fronts: the western front against traditional data protection competitors, and the eastern front with secondary storage competitors, with at least the advantage of offering a broad range of features that can allow customers to achieve savings almost immediately. It takes however a lot of effort to convince the conservative world of Enterprise IT to ditch legacy solutions once and for all, mainly because of « we’ve done it like this before » and because of the long time it takes for transition efforts to be completed.
There is therefore a real advantage for both vendors to have integration points and extend their respective ecosystem as it elevates one other’s position on the market. Snapshot integration is not an exclusive feature of Cohesity: one could envision other backup vendors to integrate similarly with Pure’s platform but the way Cohesity handles them allows for a consistently three-tier snapshot architecture: on the primary storage array, on the secondary storage platform, and finally on the cloud if necessary, all manageable from a single interface.
A second aspect is that it’s worth mentioning the fact that Pure Storage’s Flashblade is positioned as an all-flash object storage platform; usually object storage often crosses the path and overlaps with secondary storage, so there’s a real dynamic here beyond the simplistic aspects of data protection: Cohesity’s secondary storage platform can be considered thanks to these integrations as the natural secondary platform for offloading data that lived originally on a Flashblade array.
I personally like this collaboration and believe it makes sense. Not only does it reinforces the confidence of customers who already invested in one or another technology, but it also appeals to customers looking for a credible alternative to well established vendors either in the storage world or in the data protection worlds.
Even though I was invited by Gestalt IT and Pure Storage to attend Pure Accelerate (see my Tech Field Day-related posts), and had my travel/accomodation costs covered, this article is not related to content I was exposed to during Pure Accelerate and was planned to be published before the event. Likewise, it is not sponsored by any of the vendors mentioned, and my briefing with Cohesity was to get an overview of where the company is headed to and what are the latest developments in their product line.