Dell PowerMax NVMe Reviews

We are a very large customer of Dell EMC. We have several different deployments or installations. The biggest use case is probably a multi-tenant or shared environment where we provide many petabytes of storage for multiple customers who utilize that same infrastructure. We are a managed services provider in the cloud sector so we have to deliver high performance storage for thousands of customers who have to be up all the time.

There are a lot of different use cases, in general: Having large quantities of storage available that is always available, because of this uptime is important as is performance. As a service provider, we deliver storage on demand for our customers. This is important because we can adjust storage needs on a per customer basis. Whether it be increases or decreases in storage, this platform allows us to do that very easily.

We are using the latest release.

As a service provider, we have to deliver the best possible service that is backed by SLAs. The NVMe performance is fantastic for our customers and the features of the PowerMax are fantastic. We have seen improvements in performance, which means less customer support tickets. The ease of management frees up resources for our storage teams so they can focus on other problems with other platforms, etc. This is such a self-sufficient beast of a platform that it has really freed up a lot of time so they can focus on other stuff besides storage.

There is no management overhead involved in optimizing performance. It does it so well on its own. We don't have to manage much at all. It really is like a set it and forget it solution. My storage engineers love the system. It is a lot less work than our previous systems, which weren't bad by any means. There is not nearly as much management as before. So, we are saving dozens of hours per month for our storage team, and that is a real cost in our business.

There are different ways to look at security and availability. We take advantage of array level encryption, but that is a behind-the-scenes thing. We tend to focus on the availability part, because high uptime and performance are important to us. In regards to data security and availability, the data is secure if it is encrypted. The availability means that it is always up.  We have very good opinions of the security features in both single-tenant and multi-tenant deployed to the security. 

There is also the security concept regarding access to data. What we are seeing is that the PowerMax is so consistently dependable that it gives us a very solid comfort level in terms of level of trust. There is data security and protection, keeping your data from the bad guys. On the other hand, there is security knowing that your data is always available. PowerMax provides both of those.

Our primary use case for PowerMax is hosting our VMware environment with VMware SRM hosted on and connected to both. The PowerMax does the SRDF replication for VMware SRM, and some of the workload on it is for the physical environment that consists of Unix, AIX, and Sun Solaris. In addition to that, we have physical Windows and Linux servers as well. We have 1,200-plus virtual machines hosted on PowerMax.

We have two PowerMax 8000s, each deployed at a different site. The capacity of the PowerMax at the primary site is 500 terabytes, and approximately 200 terabytes at the DR site.

We are coming from the VMAX environment where the storage provisioning was a bit complex. We had to create volumes manually from the command line. But with the introduction of the PowerMax, it's a piece of cake for us. We can assign whatever volumes we want in two to three minutes. Storage provisioning has become very simple for us and is a real improvement.

The primary use case is the database and high transactional use for block storage.

It simplifies things, as we are using higher availability systems. We have always had a 70:30 effective rule between vital critical applications and those systems which are not critical, important, and discretionary. The Symmetrix's line has always been the go-to for our vital, critical application types. 

We are also implementing storage virtualization with VPLEX, where it is giving us the ability to move the storage to the proper platform based on the operating characteristics needed by that platform.

Symmetrix has always been a high-end, high availability system where we have never had one go down. I have never known one to go down. It is just an operational-stable platform that we still have to manage and maintain. However, it's not like some of our lower-end systems, like on the CLARiiON side of it or the Unity systems, where we have to be careful what we put on them and what we do with them because of the loading of the systems. 

This is why we need to virtualize the storage element, so we can move things to the proper platform. We are going to find that we will have more systems going from the mid-range to the higher range, specifically because of the needs of the platform. Wherein before it was driving by someone determining cost for what they wanted to pay for the storage, which overloaded and created problems for the lower-end tiers, because they're pulling too many IOPS out of that tier.

We have two PowerMax arrays. One is at our primary data center. The other is in the secondary data center, and they replicate back and forth to each other. We use them to store a lot of databases and files, but we don't have as much on them as we used to because our CIO is outsourcing a lot. We have taken a lot out of the data center recently, so there isn't as much on them as we intended when we bought them but I think it's mostly databases, file shares, and some one-off applications. It's all virtualized on VMware as well.

PowerMax improved our storage performance and allowed us to consolidate our old storage into one platform. It's faster than the older EMC equipment we replaced. We had a few different storage arrays, and a couple of them were approaching the end of maintenance, and one was a year away from the end of its maintenance. So it was time to either spend a ton of money on renewing maintenance or replace them. At the same time, PowerMax has made storage provisioning more difficult because it's not as intuitive as other arrays, but it's still a good solution for our mission-critical workloads.

We're using SRDF, but it hasn't affected our storage network bandwidth requirements. We haven't had any issues, so we haven't had to increase the size of any of our connections. Inside the data, there certainly wouldn't be any issues. The only problem would be replicating to the other site, and we haven't had any issues. We have a reasonably large pipe between the sites.

We are one of the largest telco companies in Pakistan. We use Dell PowerMax NVMe for telecom IT, which includes different workloads hosted on different systems. We use systems like UNIX, VMware, Windows, and Linux to map Dell PowerMax NVMe.

The solution's most valuable feature is the data reduction ratio. You can store a 5:1 ratio of your data in Dell PowerMax NVMe. For example, if you have 200 terabytes of usable capacity, you can save up to five times of data on that capacity. That means you can save one petabyte of data on just 200 terabytes of usable capacity.

The solution's end-to-end NVMe capabilities have improved our operations. Everything is fast, and you don't have to map your storage to old hybrid storage technologies. Our organization has experienced positive change, and the solution has improved our uptime, availability, performance, and capacity.

We currently use it to power our Oracle databases, especially for our core banking solution. We also use it for storage. We provisioned the storage from PowerMax for various VMs that we created for the applications in that environment.

We use the NVMe SCM storage tier feature, and that's how we're able to do the service level capability (SLA). We have storage class memory as a part of our deployment, and we have about 10% of our storage sizing allocated to storage class memory. With that, we are able to create different service levels for the disk groups or loans provisioned from this storage.

It most definitely helps in improving storage-related performance in our environment. The way our core banking solution works is that we have what we call ODS blocks. So, for leveraging that SLA, we were able to implement some kind of priority for those ODS blocks. Oracle had said that this is something for which their Exadata has a special way of doing, but based on my own assessment, we are able to achieve relatively similar levels of performance by using PowerMax.

Before we deployed this solution, we used to struggle with processing about 100,000 transactions in 10 minutes. We are now able to process about 350,000 or more transactions. These are conservative figures. We did hit much more than that, but conservatively, we are able to see about 300% performance improvement as compared to the SSD storage that we had previously from IBM. We have metrics to show that. The performance is different, and it is better than what we were used to.

We are in our ideal environment in which the storage double acts as our UAT and our test environment. So, we've seen remarkable deduplication in that environment because we are able to expand the footprint much more than what we are able to do in production. The production environment is a bit more controlled, but in our DR UAT environment, we are able to stretch those capabilities. The metrics that we see and the number of environments that we're able to create is quite remarkable.

It provides NVMe scale-out capabilities, which is pretty awesome. We currently have a plan to scale up. We started off with about 100TB. Based on the performance that we've seen, we're consolidating more workloads on the storage. We need to scale up a bit, and we find it very valuable to be able to do that. The ability to scale out and scale up marginally depending on what you want is quite valuable to us.

Our primary use case for Power Max is customer data. We host hundreds, if not thousands of applications, large and small, for hundreds of thousands of customers. It's the storage platform for our customers' online presence.

What's not on PowerMax? Because we have hundreds of thousands of customers running thousands and thousands of applications. From the small mom and pop shops running their mission-critical eCommerce site to the major Fortune 500 companies running every major database: Oracle, SQL, MySQL, Postgres, etc. We're running the big database engines, and the database is the holy grail for all online businesses. Therefore, major database applications are very important. Big eCommerce applications for very large brands are running on top of it, as well. We are running everything on it.

The role that data plays in Rackspace is two-fold because we are both a customer and a partner. As a service provider, we are monitoring millions of data points every hour for our customers. We are monitoring the health of their systems: the traditional IT monitoring, CPU, networking, storage, uptime, security, etc. We are gathering all that data and need systems where we can dump that data, then analyze it without fail, knowing that it will be there. PowerMax gives us the latency and capacity that we need at any scale for all the data that we can throw at it.

On the customer side, they are using us and our underlying PowerMax infrastructure for their mission-critical applications to do things with big data, dedupe, and other applications. Our customers are using us for the foundation of their big analytics applications.

There are a lot of ways that PowerMax is helping our organization function. From a storage admin standpoint, there is no longer a need to rely on the command line interface (CLI) to get data which is needed for performance monitoring and troubleshooting. When you use a CLI, you're actually requesting copies of data and impacting the performance of that production's data. With the REST API, we can do things, just ask for it, and there is the info. We are not impacting production systems.

We are using the PowerMax for our core banking solution, ERP, and our payment systems, as well as middleware, ATM, and the most critical banking systems.

The main improvement for us is that we have seen up to 12x performance improvements after moving from earlier, mid-range Dell EMC storage to PowerMax. Some of our reports, which were long-running, are now completed in a few minutes. Something that would take two hours is completed in 15 minutes and that has improved productivity. 

We also used to get timeouts from our storage, but now, after migrating to the PowerMax, there are no timeouts because the latencies are in microseconds, compared to the milliseconds of our old solution.

Our bandwidth requirements have been reduced because of the compression and the dedupe that we are getting with the built-in SRDF. It is bandwidth-optimized. And the best part is the reverse replication. Suppose you activate your DR. When you have to come back to the main array, only the changes are synced. That is unlike many other products. Here, only the changed tracks need to be updated, making the reverse replication very fast.

Also, by enabling the compression and deduplication, we get a very good level of compression and dedupe, of 4:1, which means if you have 40 terabytes, you only need to buy 10 terabytes. There are cost savings there. And by default, thin provisioning is in place, which also gives you at least a 40 percent reduction. And because of the bandwidth optimization, the link required for the DR replication is also reduced, meaning you are saving on the bandwidth costs. We have easily saved 50 percent.

Overall, you are getting very high-performing and reliable storage.