Pure Storage FlashArray Reviews

We use it for performance, the capacity of deduplication, and compression of the data.

We are a small cloud service provider. When we put Pure Storage working on the services of our customers, most of the problems that we had in the past with the performance in IOPS have disappeared. It has been a great improvement for our customers' services.

The storage is very simple. 

We can use more capacity because of the compression and deduplication.

The predictive performance analytics are good, easy to use, and simple to see. They are simple to understand, not complicated.

Going forward, don't complicate things for the customers. 

The solution is very stable.

Only one disk has a problem. The performance with that problem doesn't create problems for our customers. We are able to maintain the performance of the program.

We have already made upgrades now for two months. We think that the scalability is very good. If you want to go to another array or add more capacity, they will change it, if you have the support. So, we put more capacity on it. There is a simple way to do it which has a protection of investment.

We have only used the remote technical support in the case of the disk. They are very good. They acknowledged the problem quickly, identified it, and are always asking and seeing things, which in some cases, are more difficult for us to see.

Because of our clients, we needed a more structured solution with performance which was stable. So, we tested new storage, and Pure Storage was the one that revealed to be more flexible and simpler. 

The initial setup was very simple and straightforward.

For the integration, we used a reseller. It was so easy to put in place and put it to work. They did a good job, but I think we could also do it.

For us, as the customer, it reduced the price of the management. The total cost of ownership has been reduced. In the beginning, the investment was greater. Now, it is about a 20% reduction.

Using older techniques, we see that we can offer clients more capacity. The real capacity that we get to customers is six or seven times greater than the capacity that we have in place.

In the beginning, we saw that the price is not very good. When we made some compilations about the deduplication and the compression and what the equipment does, including the differentiation of upper management of the storage, the price was not so bad. However, in the beginning, the price was very difficult to justify.

It is a very good solution. It responds to all the workload problems that we have. It could be with some different workloads that the solution might not respond the way that it responds to us. Test it. People will be astonished with it performance and simplicity.

We have two arrays in two data centers. Normally, in the arrays, the latency is about 4.3 milliseconds per second, which is very good in all workloads. In terms of reduction, our customers are seeing about a seven to eight plus reduction in the capacity that they have.

The TCO for flash and SSD implementation are comparable.

We use it for nearline storage.

Right now, we just have lab equipment that we test them on and we try to compare them with other solutions.

The thickness and the sizing for when we put it in the data center. Also, the performance and price.

It has good stability for our company.

It's granular.

The support is good.

Our storage is old, so we were searching for what would be the next good solution for us. We had an integrated solution with a supplier, so we were looking to get rid of this kind of model. 

The first set up we had was really straight forward and simple.

We used a retailer to buy it and it was easy.

Compuverde. But, we like to have data sheets and a more traditional storage than a complex unit.

I would rate this a seven out of ten because it's a good performance storage, but the price is a little bit high. Our predicted performance analytics is also going really well, so if you need faster storage and a good product, this is the one you should go ahead with.

Our primary use case of this solution is for Rack Database Storage and Virtualized Server Storage.

I used to have to manually sync storage from server to server, for multiple clusters and database storage for disaster recovery but now, that's all automated. I set it up once, and it was done on the very first day we implemented bolts. Storage arrays were set up on the very same day and by that afternoon all of the replication was configured and I haven't had to touch it since.

Has also helped simplify storage for us. It had taken the original person we used a full week to implement. With this solution, we had both arrays set up within around four hours with a thirty minute drive time between the two locations.

We have seen TCO of flash be lower than SSD implementations. We're faster which is part of the equation too. We're paying for speed and if we would have had to buy other solutions, then that would have cost us on the morale side and on the user satisfaction side.

Another way that it has helped my organization is that now we are seeing 3.6 to 1 on Oracle Databases. Our goal was 3, we had to have 3, so we got 3.6 to 1.

The most valuable features for us would be the ease of implementation and box-to-box replication.

I would like to see box-to-box encryption on replication included in the next release. 

We've had zero drive failures and zero problems with it. We've had it in place for about a year and a half and have had zero complaints, other than that box-to-box replication is not encrypted.

Scalability appears to be easy but we haven't had to use it. 

We had to open a ticket with their technical support. We needed to get NIST certified and we had some of our storage on that PRA and we had to be done at the end of the year 2017 so we were in constant contact with support to ensure that we were going to meet all the requirements. In the end, we did hit that date.

We knew that we needed to switch solutions because we were using very old Dell EMC hardware. It was painful, we had weekly drive failures. Every single week one of the key drives failed. It was old, it was out of support, we were losing support, we were paying for extended support, we knew that we had to have this solution. It was all spinning discs, there were a couple SSDs on there but for the most part, it was all spinning discs. We saw some major improvements.

The initial setup was straightforward. The other guy we used took a week to set up and there are still issues. Here we had two arrays set up within four hours with a thirty minute drive time.

We used a reseller called Sirius for the implementation. They were good, we didn't have any complaints. 

We looked at Hitachi which we did put in place for some of our dev environments. We also spoke to IBM. We used to use Texas Memory Systems which was bought out by IBM and we reached out to them to see if there was an equivalent and there wasn't. 

I would rate this solution a ten. 

The features of this product that are most valuable include the data reduction capabilities. Being able to write and read data at incredibly high rates of speed and reduce the footprint of your data by three to four times at least is invaluable. Also the ability to clone and snapshot the data with little to no additional size impact. In the database world this function has quickly become a must have and not just a nice option.

The organization in which the product is deployed had many instances of the cache databases. There were instances in production, backup, test, development and others. There were then copies of copies. There were updates to databases and refreshes, as well as exports for reporting. There were more than 20 instances of the same data that existed in production. The space reduction and cloning capabilities alone resulted in major cost and administrative overhead reductions across the board. That doesn’t even account for the ability to reduce RPO/RTO to very low expected time frames.

The overall scalability for this product could be improved as well as having a single console to management multiple arrays. The scale is constantly being addressed as SSD drives get larger so will the overall capabilities of the arrays.

I have used this solution for less than a year from POC to production deployment. [Pure Storage FA-420 and (2)FA450 array.]

There were no issues during the initial deployment of the array.

There have been no issues at all with stability. The only issue was on our side, procedurally - how we did our zoning needed to be modified. Instead of zoning in groups of hosts we needed to do individual zones for each host to the arrays separately. This had nothing to do with the array though.

There have been no issues at this point. There has not been a need to scale yet, however the scale prior to deployment was more than sufficient.

Customer service has been outstanding. During the HA testing in the POC period we had people contacting us to verify that status of the array before we had completed the testing phase. They are thorough and capable.

Implementation was flawless and any remote support they handled any questions we had with ease. There really haven’t been issues of any nature that required much technical savvy though.

We had not previously used an all flash array. We POC’d the next best competitor, but feature wise, overall system stability and space reduction numbers made it so that Pure was the clear leader.

It took about an hour to get multiple arrays installed and deployed. They were incredibly easy to setup and even easier to get online and provision storage.

We implemented through a vendor, but the manufacturer did the actual installation. There were no PS hours incurred for the installation from the vendor.

We evaluated through technology research: Whiptail, SolidFire, XtremIO and Pure Storage. Ultimately we POC’d XtremIO and Pure Storage.

Definitely take the time to perform a POC. There are multiple players in this market right now that have good offerings and specific environments are very different. Also, don’t get sucked into the IO race. At the end of the day, the number of thousands of IO didn’t really have that much impact on our product, because minimum IO reqs were far below anything, any of the arrays advertise and produce. Also, just because an array can do a million IOPs doesn’t mean it is the right fit for your environment.

This storage has to be the most well thought out and effective storage that I have had the privilege to work with.

We are a system integrator and this is one of the solutions that we provide for our clients.

For our most recent customer, this solution is being used to host VMware workloads in an on-premises deployment.

Our customer has been able to migrate some of their cloud services back on-premises, which is of benefit because they were having some performance issues in the cloud.

The most valuable feature of this solution is its simplicity.

It is easy to deploy and it's all-flash, so it's very fast.

As partners, we should have the option to download the software, rather than have to go back through Pure to obtain it.

This solution seems to be fairly reliable. I haven't had any issues personally, or outages or anything of that nature.

The flash array scales up well, but it doesn't really scale-out. I think that's more where Pure's FlashBlade comes into play. You'd have to replace it with the next model up controller, in order to scale.

I haven't had much to do with technical support. Typically, we'll deploy it for our customer and hand it over, and then they'll manage it from there.

Prior to this solution, our customer used Dell EMC, but there were several problems. They were not happy with it because of the performance of the array, and also the complexities around engaging the support teams. Their solution had also reached end-of-life, so they had to move to a new platform anyway. Ultimately, they wanted to move to something simpler than their existing platform, so they moved to Pure.

This initial setup and deployment are pretty straightforward.

For the setup, it is very simple to populate the information. We had a workbook that was provided by Pure to give to the customer to gather all of the details. That made the implementation very straightforward.

The one thing that was a little bit annoying was the fact that we still had to come back through Pure to get the software. A lot of our other partners, we can just download it ourselves, and as long as we've got the process, we can go and do it ourselves. But, in this scenario, we still had to come back through Pure to actually get the software.

We implement this solution for our customers. We are a partner and we're certified to do deployments.

I think that our customer has seen ROI because their existing solution was getting into extended maintenance, so it was costing a lot of money for that service. Also, with less time spent managing that old array, they had more time to do other things.

In addition to this solution, our customer considered another Dell EMC option, as well as one by Hitachi. They chose this because of the simplicity, and the fact that it is all-flash meant that they didn't have to worry about performance. 

VMware benefits our IT organization because we are partners, so we deploy VMware services. It also helps our customers make their applications more readily available and reliable.

We are using the VMware plug-in for Pure. It's meant more rapid provisioning of volumes for VMware, and it gives the customer more visibility of the storage.

Both the ease of setup and the reliability of the array makes it quite simple to manage for the customer.

My advice to anybody who is researching this solution is to consider that things are changing a lot in the industry at the moment. So, obviously, looking for things that are going to take less time to manage, are easy to implement, and give a good return on investment are important success criteria.

I would rate this solution a nine out of ten. 

My rating of Pure Storage is a ten out of ten because of the price for performance and footprint - the overall value.

There's always an opportunity for new feature functionality. It's just a question of what that will be and what does the future look like?

It's very much a stable product.

It is absolutely scalable to our company's needs.

We've used tech support and we have found it to be very favorable.

The implementation process was seamless. It was very quick. It took less than a week to deploy.

The return on investment is good, very strong.

In comparison to the competitors, Pure is very price-competitive for the future functionality that it provides.

We considered Dell EMC, we looked at Nutanix, Cohesity, IBM, HPE. We ran quite a bit of the gamut.

I would absolutely recommend this product to a colleague. We have no regrets.

We have one person who administers it. We have about 9,000 employees and the IT department has about 300.

VMware is currently our main use case because it dedupes really well.

Because we were able to afford to go all flash, we don't manage tiers, we're not moving data up, and we're not waiting for overnight cycles. 

Firstly, dedupe is the most valuable feature. Hands down. Simplifying storage is also a big win overall. As far as the monitoring with the latency goes, we're not monitoring the apps to see how they're doing at different periods, which saves us time. We're just setting thresholds, walking away, and waiting for emails if they're broken.

The big thing would be to simplify the compatibility to Openstack. The Openstack going into Nova works really well, but if Pure had a few more of those features that would be my win.

It's been rock solid.

So far it seems good.

We've used them a few times, mainly with upgrades. They are quite responsive and our local team was pretty good.

They gave us the rundown and was simpler than expected. They gave us the sheet and sent us off saying, "Hey, if you need us, give us a call." We just plugged it in and up it came.

We used a reseller, but then we did our own deploy along with their help. They came in and gave us a training course so that we could maintain it ourselves.

There are a lot of companies that give a solid performance and a lot of places you can get flash. The pricing wasn't that much different, It's really the simplicity that makes a difference. If the data starts flowing too fast, it slows things down and does it later. Those features are the winners for us.

We're constantly on the hunt, and we always keep three to four vendors in. Usually, it's been the bigger players, the IBM's, the EMC's. They all have their strengths and weaknesses, but we were looking for something a little different this time around.

We use Pure Storage on our databases. We have massive SQL databases, four-node clusters and we present a LUN directly to them. Then we have Fusion-io cards as a backup. We also use Pure in our data centers to replicate our databases for our DR center so that we can be secure. 

Pure Storage has helped improve our organization because before them we had a 3PAR of a giant V400 and every day we would lose a disc or a magazine. We had to call out a guy to come onsite. It was a massive three-rack thing. Pure Storage, it's really modular, we're maxing out shelves where we can, and it doesn't take up as much space, it's not as hot, its a lot better than 3PAR.

Replication is the main reason we have it. It has helped to simplify our storage in the way that it just simplifies and there's nothing to really set up. Once we have them linked we ship them over and we sit our RTOs and our RPOs.

As dedupe and compression go up and we get more out of it, then we do see reduction in total cost of ownership. We're also throwing more and more on than we ever had before, so it's hard to tell, but we're getting more data on a smaller array than we ever had before.

The 3PAR SSD arrays that we have are still failing a lot so even though we're under warranty, we still have to get someone out and usually have someone troubleshoot so that usually adds onto the cost. With Pure, we've had a disc fail and we pop it out and you pop it in and it's good to go.

In terms of performance metrics, depending on what we have on it, some of our databases will get 4.8:1. When we do a big release our SQL tables change values so we'll see that reduced and we'll go up to sometimes 110% utilization. We're working with Pure Storage to try to fix that and see what we're changing so much. We also mistakenly had a 10pb on Pure so that data churn really reduced our usable storage. We're learning how to use Pure properly.

The magic that the storage does would be the most valuable feature for us. Deduping on the fly is really cool to us because some of our stuff we get around seven to one, which is amazing. I definitely like the new redesign of the UI that was done. We just had to do a DR test, and we had to make snapshots and copy them over, and it was a lot easier to use I think with this new UI than the old stuff.

Stability is good, we haven't had any issues. The only thing is that we've had to upgrade controllers a few times because we ended up wanting to use more stuff on here. At first, just our databases, then we moved our VMs to it. We really haven't had any issues except just needing to upgrade to bigger controllers.

We stream into StatsD from Pure Storage, LogicMoniter, and a few others so we don't use the UI performance manager as much because we like a single pane of glass but it's got everything I need. When we do see latency or we have issues it's usually really clear from the graphs.

In terms of scalability, we buy new controllers or we build new shelves and we're able to scale out pretty much whenever we want, as long as we have the money to spend.

We will usually hit up technical support for something that's not too major. We've never had a SEV1 outage with Pure but we've enabled remote support. They log in and they're good to go.

We switched because we were running out of support with 3PAR and they wouldn't renew our support unless we got a new array which was a lot of money. We had some of those SSD arrays, we didn't want to put all our eggs in one basket so we spread the vendors by having an SSD array from HPE and Pure. Once we solved the data reduction and what Pure does we were hooked. 

The initial setup was very easy. 

We did the implementation and worked with professional services. For the most part, our main guy in the compute team has had experience and it was pretty simple. We didn't need a forklift like we needed for the HPE. Just rack and sack and ready to go.

We have seen ROI just from being able to move our databases around, because we have different pods, quickly and specifically. With 3PAR we'd have a lot of remote copy failures, and that doesn't look good for an audit or for a DR test. We haven't had any of those problems with Pure, so we spend less time troubleshooting.

We have a bunch of different storage, like Isilon from Dell EMC, NetApp, HPE 3PAR, Cohesity, and Pure Storage. They're all different functions, and Pure is our warrior, if we need something really fast, really low latency.

I would rate this solution a nine.

If someone was considering this solution I'd definitely ask them what their use case for was. If they had a high workload, like for example, I have a buddy who works in the entertainment industry, and they need to edit 4K video, so they need something like Pure that's really fast. I love the support and I love just what Pure does in general, so I'd definitely suggest it.