Jason Michaud, CIO of MacStadium, an infrastructure-as-a-service provider focused on Macintosh-based software development environments shares his perspective as a customer of FlashArray and a beta user of FlashArray//X, the First Mainstream 100% NVMe All-Flash Array. This blog is Part 5 in our FlashArray//X launch series:
Part 5: A Customer’s Perspective – How MacStadium Uses NVMe [this blog]
Every day, customers come to us seeking higher performance, whether it’s for application development, building the next communications platform or providing macOS VM solutions. We understand how critical software developers are to meeting time-to-market deadlines and why our customers have an urgent need for their teams to be writing code, not waiting for an IT infrastructure to respond. This is why MacStadium exists. And it’s why we use Pure Storage to help address those needs.
Since installing our first FlashArray//M in early 2016, we have been able to give our customers an unprecedented level of performance, reliability and scalability. Customers leveraging our hosted Mac private cloud tell us that our Pure-based “Performance Tier” level of service provides system response times equal to dedicated desktop workstations, with the added benefit of scaling their operations easily and quickly, without any loss of performance or efficiency. That is a huge benefit to our customers, and it is a big reason our business has continued to grow rapidly.
While scalability is key, it’s just one of the many benefits we receive from Pure Storage. We have been greatly impressed with the high data-reduction rates, consistent performance, ease of deployment, and the effortless management of Pure arrays. We now have a dozen FlashArray//M systems, and many of them have been upgraded to meet growing customer demand. Every one of those upgrades has been carried out flawlessly during production hours, without any downtime or loss of performance.
Given this track record, we look forward to upgrading our //M arrays to the new 100 percent NVMe All-Flash Array, FlashArray//X. This upgrade is also non-disruptive and will be fully covered by our subscription to the Evergreen™ Storage program.
As a beta customer deploying FlashArray//X in a rapidly changing IaaS environment, we require a modern cloud storage solution that offers us the greatest flexibility and the most cost-effective growth path to sustain and expand our business operations. First and foremost, it must be exceptionally simple to manage and based on open standards, so we’re not locked into proprietary systems. And it goes without saying that it must provide elastic scale to meet on-demand capacity and performance, as well as multi-workload sharing required by our customers. Plus, it must be “big and fast” – meaning hundreds of petabytes of capacity and tens of GB/s of bandwidth with consistent sub-millisecond latency. Finally, it’s got to support fast networks that deliver high-speed Gigabit Ethernet-level performance as well as PCIe, NVMe and NVMe/F technologies.
In our testing, we have seen that the //X delivers notably faster performance in the same footprint as the //M, meaning we get a lot more I/O for the dollar.
The real test, though, is does it have an impact on our customers? And the answer is a resounding “Yes.” Customers using services powered by //X arrays report less “back pressure,” which means latencies are lower, bandwidth is higher, and when demand spikes, response times are not impacted.
If you think about the software development process as an assembly line, the //X makes that assembly line function faster and more predictably. Based on the comprehensive testing we’ve performed to date, the //X is a huge step forward, for both us and our customers. Any of our customers who have transacted recently with MacStadium have likely benefited from //X. From our perspective, Pure is redefining flash storage for the cloud era.
We thank Jason for his participation in the FlashArray//X beta program and for sharing his insights on this blog.