Pure Storage recently announced a new offering running as software on AWS known as Pure Storage Cloud Data Services. So what does a cloud offering from Pure Storage get you other than an Orange Colored Sky? A lot of value to the enterprise, as it turns out.
What Is It?
Pure’s Cloud Data Services consist of three distinct, but complimentary offerings: Cloud Block Store, CloudSnap, and StorReduce.
Cloud Block Store is Pure’s own Purity OS running natively on AWS. By standing up an EC2 instance back by AWS Elastic Block Store (EBS), Cloud Block Store can present an iSCSI LUN back to other EC2 instances or even VMware Cloud on AWS if the customer is using the service. This is significant because it allows customers to leverage services previously unavailable in native AWS storage services. Space saving features such as deduplication, compression, and thin provisioning will help keep costs down. Active clustering within a single Availability Zone or across Availability Zones within the same region will deliver High Availability. Further feature lists and use cases for Cloud Block Store are detailed in the deep-dive blog on the Pure Storage website, but suffice to say the offering brings Pure’s enterprise grade storage to the public cloud.
CloudSnap, the second announced service, will allow existing Pure Storage customers to leverage the public cloud as location for Pure’s portable snapshots. While the snapshots are not a new feature of Pure’s FlashArray, the ability to send them to an S3 bucket is. Not only will this allow customers to shift storage of snapshots off site as a viable long term storage strategy, but it will also allow customers to mount snapshots to a Cloud Block Store for enhance recovery capabilities.
StorReduce is not actually an entirely new product. Pure Storage announced the acquisition of StorReduce, the company in August of 2018. The announcement of StorReduce for AWS is the end result of that acquisition. When running on premises, StorReduce software is backed by Pure’s FlashBlade product and allows for quick recovery of backups on premises. The power of StorReduce for AWS is unleashed when a deduplicated instance of the customers data is replicated into AWS S3 allowing for long term retention and native recovery of data in the cloud. An excellent use case for StorReduce would be to leverage it with a backup platform that does not natively support cloud storage as a means of sending data offsite and providing extended retention.
Why Would the Enterprise Care?
The Hybrid Cloud is increasingly becoming the reality in the enterprise. The decision is no longer being made between only leveraging private cloud vs public cloud for an organization as a whole, but rather on an application level basis. But moving data in to the cloud isn’t as simple as signing up for an account and pushing data into an S3 bucket. The enterprise is full of applications that don’t fit the definition of cloud native. They may not natively support cloud storage, or expect a certain level of high availability to be delivered by the infrastructure rather than handle such capabilities inherently.
While it may be possible to rewrite/refactor such applications to fit a cloud native architecture, such an undertaking may be impractical. While an application may be critical, it could have a limited lifespan in favor of a new application currently in development, so why bother rewriting it when the plan is to eventually transition away. Or what about an application that is heavily used, but was not developed in house? How is an enterprise supposed to refactor for the cloud aside from submitting multiple feature requests to the vendor and waiting for them to modernize?
There may even be a situation in which an enterprise needs to stretch an application across their on premises infrastructure and the public cloud, make it a truly hybrid cloud application. Wouldn’t it be simpler for the developers building and maintain that application to use the same set of tools and APIs to build and maintain such an application?
Being able to extend an application into the cloud without having to make significant modifications is highly valuable to an enterprise. It is becoming more common for an enterprise to invest in a service that fits their need as an alternative to building more infrastructure. In fact, many organizations are interested in shrinking their datacenter footprint. Such an edict coming down from on high means that there isn’t a single “right way” to adopt cloud for the enterprise. Leverage native and third party services to achieve the desired outcome is the way that enterprise does cloud.
Ken’s Conclusion
In a couple of years, we as an industry are likely to see 2018 as the year public cloud grew up. What was previously the domain of startups and cloud first development houses has turned a corner to become enterprise friendly. Whether it is a public cloud service provider releasing new features or products or a third party creating new offerings that enhance existing cloud offerings, we are seeing increasing relevancy to the enterprise.
Being able to leverage the cloud’s self service capabilities as well as a consumption based pricing model are huge to the enterprise. Not requiring an enterprise to refactor applications or wait for a vendor to do so before being able to reap these benefits means both a more rapid pace of the adoption from cloud holdouts due to the onramp being much smoother than it was previously.
By releasing these three products, Cloud Block Store, CloudSnap, and StorReduce, Pure has done more than produce a me too product that we have seen out of many traditional enterprise infrastructure companies. Pure has provided the tools necessary for their customers to overcome the inertia of an old business model that has prevented many enterprise customers from adopting cloud. I expect to see an accelerated pace of cloud adoption from Pure’s customers who choose to leverage these tools.