Don't be fooled by their marketing. They aren't BS-ing you, for a change. Try it out and check the TCO for the next decade.
September 18, 2019
Don't be fooled by their marketing. They aren't BS-ing you, for a change. Try it out and check the TCO for the next decade.
Score 10 out of 10
Vetted Review
Verified User
Overall Satisfaction with Pure Storage FlashArray
We use Pure Storage Flash Array X and M for our primary and data recovery sites. Pure Storage is our primary storage mechanism for an enterprise workload that has a healthy mix of applications, databases, VMs and file systems. Deploying Pure Storage not only addressed our performance, cost, manageability, flexibility and investment protection (hardware support lifecycle) problems but also has introduced new avenues for us in supporting our business units to make decisions easily about dynamic business plans that explore IT systems' ability to support those plans with not much overhead - both people and money-wise.
- For a change, it's the first storage system I've used in about 20 years that works as it is written on the brochure. In fact, it over delivers. We have had several bad experiences in many ways with storage systems and when you speak about Pure Storage from their datasheets or any other material, it feels like they are going overboard with what they say. But then, only when you deploy it completely would you realize that in fact, at least some of what they mention is underrated and the system is not marketing BS.
- The support is awesome. In addition, we could never imagine upgrading a production storage system during business hours without missing a beat and without dropping any sweat. In fact, you don't need to do anything but log a call with support 2 days in advance and allow them to provide remote assistance. They just do it for you and no downtime or loss in performance. We tried it twice and it still feels unreal.
- In addition, when we didn't get their promised data reduction at our DR site initially, by the time we discussed internally and called them to check what to do, they told us that they had already assessed the situation, made a report for it and submitted it internally for upgrading our disk capacity free of cost. No words. Just great. By the way, funnily enough, after they added capacity, we started getting far more data reduction than they promised. Not sure why it acted up initially.
- Performance. Be it - IOPS, latency, redundancy when disks/controller/network cables fail (yes, we yanked everything to make it fail, and it was fun) it doesn't fail. Even the unreal looking data reduction, they nailed it.
- They said it takes maybe 30 minutes to learn their system and no training is needed. We laughed at them. Not once, but many times. I don't think it took us any more than 30 minutes. This is about the easiest system we manage in our datacenter. Almost zero learning curve.
- There aren't many gaps and they've been closing those few very quickly. An NFS/SMB server without having to worry about MS licenses would be great.
- More VM based usage analytics. We understand the technology limitations in getting the details when using VMFS volumes instead of NFS volumes but it can be slightly better.
- Out business projects have accelerated like never before. We have great confidence that the storage system is modular (thereby not making us spend huge amounts of money for each expansion) and that our investment is protected well in the long run due to the very long support life of the products while keeping the maintenance cost filed/flat.
- Our virtualization environment is now running so much smoother and we reduced almost 20% of compute capacity/resource utilization right after we moved datastores to Pure FlashArray.
- We moved from 2 hybrid storage systems of half a rack each to 3U with half capacity filled, reduced latency 10-15 times (depending on which application and which storage system the application moved from), reduced purchase costs and maintenance costs by at least 25% compared to alternatives evaluated.
- As a negative impact, the IT infrastructure team is now dealing with more business projects and is much busier delivering those projects instead of troubleshooting performance issues or attending constant meetings explaining why we can't take on a project or calculating how much investment is needed for infrastructure for a new project.
We considered Nutanix HyperConverged systems, EMC XtremIO, NetApp SolidFire; evaluated HP Nimble All Flash & Hybrid, NetApp AFF series and VMware vSAN. The former 3 were ruled out for being too expensive and in some cases offering too little for too much money. The later did give a tough fight on paper but once you see the system, the choice was clear. HP Nimble too made a lot of sense and was also a little cheaper to buy but not TCO wise.