Dell PowerScale Reviews

TGen is a nonprofit biomedical research institute. Our focus is primarily on genomics, translating discoveries in the field of genomics into treatments for patients.

It is central to our data storage of scientific data. We sequence the human genomes of folks with different diseases, primarily cancer but also other disorders, e.g., rare childhood disorders and people with mitochondrial diseases as well as neurological diseases. When you do this, it generates a considerable amount of data. Each time that a whole genome sequence is run, you generate anywhere from four to eight terabytes of data. For example, if you are looking at 1,000 patients, that can be anywhere from four to eight petabytes of data. TGen has about seven petabytes of storage being used for storing these genomes, which is a fair amount.

Isilon is an on-prem, scale-out storage. The nodes are linked together through a back-end high-speed interconnect. 

We are running current versions of software on the node. It has versions now. The nomenclature is sometimes not the easiest to follow, because they still like to rebrand things. 

It has given us the capability to focus on our prime objective, which is science, without having to necessarily be concerned about the back-end infrastructure that powers it. This is something we are always looking to achieve: Being able to focus on our prime mission without having technology get in the way. Scientists don't want to learn all about your storage system. They just want to do their science.

It is a critical piece for storing scientific data for our Institute. It is where we put our most valuable and precious data. We also leverage it for work on administrative data, spreadsheets, Word documents, etc. So, it is flexible. We access it via NFS and SMB. Those are the two primary methods of access that we use along with some others, such as S3 for some particular use cases.

Deploying and managing storage at a petabyte scale using Isilon is extremely simple. The user interface for management tasks is intuitive. The documentation is thorough and good, and if you get stuck, then the support is very capable. Overall, I have confidence that we can easily deploy, manage, and maintain systems without needing a huge amount of expertise to facilitate them.

PowerScale has helped us by consolidating the data without having it dispersed. Prior to this solution, we would have many different physically separate storage solutions. To do the science, sometimes data needs to go from one place to another. Moving your data at a petabyte scale, or even at hundreds of terabytes, is very time-consuming and expensive. By having the consolidation within these clusters, it has enabled us to very easily access and compute data without having to push it around to a bunch of different places.

We have a "thinly provisioned" workforce. One of the crucial aspects is that we can continue to scale a solution without having to add more humans to take care of it.

There is a reason that we chose this platform to store this priceless data. We know it is resilient. It also provides data protection that helps me sleep at night. 

One of the most important factors about it is you can manage a lot of storage without a lot of people. Therefore, ease of management is really important for us because we are a nonprofit. We don't have a huge IT staff to support a pretty substantial IT infrastructure. So, ease of management is always a really crucial consideration.

Another aspect of the management that is super important is having the CloudIQ feature to monitor performance and other data remotely. We have four clusters that we manage. Having all those clusters, being able to have a single dashboard to take a look at the health of everything every morning, helps out a lot.

One of the nice things is that they have several different node types spread all the way from super high performance, flash-based storage nodes through more of what we consider an archive tier. So, we are able to use technologies, what Dell EMC has labeled SmartPools that will tier data automatically between different types of storage. So, we can ensure that hot data resides on the high-performance storage. Whereas, once data has gotten colder, then it can be pushed off to the low-performance storage to help control costs.

We have used the solution’s support for the S3 protocol, but in a limited use case. We are looking to expand that because we are doing more work towards cloud-based solutions. So, having the flexibility of S3 is important as we design new workloads that will be more cloud-centric. They will be able to use that protocol to access data on nodes without necessarily having to go back and refactor everything.

It is good and efficient when maximizing storage utilization. The operating system behind it, called OneFS, provides granularity, data protection, and control. So, you can actually adjust the amount of overhead being consumed for your data protection, depending upon what your needs are. It is pretty efficient at keeping data protected. At the end of the day, that is one of the most important things: Knowing that your data is safe.

Dell EMC keeps adding more features to the solution’s OneFS operating system. We have been iterating with them for quite some time. The solution is continually improving and becoming more robust and reliable. One of the latest things that really helped us out was the ability to perform upgrades without having cluster-wide outages, which is huge because we don't want to shut down operations unless we absolutely have to. Having that was a really big win for us. This saved us time. More importantly, it has kept our labs functioning during upgrades, as opposed to having shut down sequencers for a day while we go through and upgrade everything, which is important.

Something that still could be improved upon is adding additional node types of different sizes to facilitate a better way to run in distributed offices. For example, we have a lab up in Flagstaff, but they don't have a lot of IT infrastructure. Therefore, it is not really appropriate to run this system at their location. So, we run it down here in Phoenix. It would be nice if there was a smaller solution that we could deploy up there that was still as cost-effective as the bigger solutions.

The thing that they are working on now, and we are following closely is more native cloud integrations. The way that we envision workloads in the future is around moving compute to data instead of the other way around. So, we would like to have a single pane glass to manage storage across a variety of different platforms, including native cloud. That would be awesome.

We were using PowerScale before Dell EMC even bought Isilon. So, we have been using it for some time now. 

We have run this product for so many years now. I can count on one hand the number of times where we have had any kind of issue that impacted availability. Usually, it turned out not to be the cluster but something else. It is extremely robust and continues to function.

We are not super aggressive in patching or anything. We believe that stability is number one. Availability is just of the most critical importance so that is really where we focus.

Once you have set up your initial cluster, adding more capacity to it is extremely easy. It is so easy that one of our salespeople added a node to the cluster. Having a salesperson do something technical is always a little bit interesting, but they didn't have any problems at all. "Boom," and it works. 

This is one of the nice things that goes back to that whole ease of management. Being able to add additional capacity is pretty simple. You just buy the nodes and plug them in, as long as you have enough of the right kind of node types. However, if you meet all that criteria, it is that easy to do.

Since it can scale so easily, as long as I have money to buy more nodes, I can grow it as big as I need to. That is important in our business. As sequencing technologies continue to evolve, and as those technologies evolve, the amount of data generation never gets smaller. It just always seems to get bigger. This is one of the absolute key aspects: We can grow on demand without having to forklift stuff. 

I have done forklifting, and it is a drag. I don't want to do that again. We want to just keep being able to grow as we need to ensure our customers have the resources that they need to do their work.

I have worked pretty closely with their engineers over a number of years. They have implemented several different items that we have suggested.

The technical support is excellent. They have good support teams within Dell EMC, but also the VARs that we use have been extremely good at helping us as well. We kind of have multiple different angles of support, and that is one of the reasons that we continue to invest in Dell EMC. They have a model that we can rely on for getting the right answers.

I would rate the technical support as a nine out of 10, because nobody is perfect.

We got our first cluster in 2008. Before that, we were using JBODs connected to Linux hosts. This was a homegrown solution. Frankly, there wasn't really anything available at that time that could meet our needs which didn't cost millions of dollars. So, we went from something that was good enough to something that was much better.

We switched because we needed something that scaled much larger than what we could build and comfortably support. That was the number one reason. Number two was, at that time, I was still doing all the technical work, and I was the one building it. I had too many other things to do. So, I needed to find something that could be supported by other people, not just me. This was really getting something that we could run in a more enterprise-type fashion, as opposed to something that we built because we had to and there weren't any other options.

Today, we have two individuals responsible for storage. Not just this storage, but any other storage systems that exist. Previously, while the storage was a lot smaller, it still took about four of us working on it. By having a single platform, where we can run a variety of workloads on it, this enabled us to not have to continually grow our storage administration staff, even though our data footprint increased many fold over the years.

It was straightforward. There wasn't anything super complex about it. 

We just deployed a new cluster last year. It took around three to four months before it was really cranking in full production. Once they are running in full production, they are adding value.

Even to this day, if we still run into something that we are not sure about, we can call support or get local support, who generally get things addressed quickly and to our satisfaction.

Since I have to manage all the budgets, I always want things to be less expensive. However, I would say the pricing is fair. Their costs are in alignment with their competitors. It is a good value for the money.

Like anything else, it could always be less expensive. That would be great. At the same time, I would like to make sure that they keep innovating.

We went pretty much straight to the Isilon product. At the time, there were no other products available that did what that product did. They were kind of unique.

We keep going back to them even though there are other products now that report to have similar characteristics. We keep going back to them because it has been such a good experience. We have a high degree of confidence in Dell EMC being able to deliver a product that meets our needs. It is cost-effective and helps me sleep at night because a lot of the data is precious. Sometimes, you get samples that you would never be able to get again, where they are kind of a one-off thing. If you lose them, then they are gone forever. We have to bear that in mind. That is really why we continue to invest in this solution.

I would rate it as nine and a half out of 10. One of the main reasons that we have been successful as an institute is because we have back-end infrastructure, e.g., scale-out storage. This lets scientists focus on doing science, which is really important.

We're using 95% of data for user access and 5% percent for the NFS mount point. We're a startup and customer of Dell. 

It's helpful that we're able to scale the number of nodes without having to build additional clusters. We started with a very small footprint and now we have 30 nodes and recently expanded an additional eight nodes on the cluster. We can create multiple storage pools from this if we decide to add a location within the cluster itself.

We're struggling to find the NIXI protocol. It's for people needing to access using Windows and Linux. We're struggling with the UID mapping and how to configure mapping-related things. I'm looking at how to map those GIDs and UIDs. 

I've been using this solution for four years. 

The solution is stable. If it's being used for the NAS protocol, it's very stable.

The solution is very scalable. 

We have direct Dell support only.

The initial setup is straightforward. We have 4,000 users in the company who are accessing the shared drive without any problems. Maintenance can be done by one person. 

We have a five-year contract with Dell. We get new hardware each time we renew the contract and the cost is calculated on a percentage-wise and scalability basis. Every five years, we replace the tech nodes. 

If you're looking for a product to use for an assembly protocol, this is the best solution on the market.

I rate this product nine out of 10. 

The most valuable feature of Dell PowerScale (Isilon) integration with other solutions because of the standard file system protocol.

Dell PowerScale (Isilon) could improve the load distribution capability. For example, in some cases, the system load is not distributed automatically on all the nodes but is concentrated only on one. You have a peak request on only one node and the others don't do anything.

In an upcoming release, the solution should have security features embedded, not external software.

I have been using Dell PowerScale (Isilon) for approximately nine years.

Dell PowerScale (Isilon) is a reliable solution.

Dell PowerScale (Isilon) is scalable. It is easy to expand capacity.

Most of our customers are enterprise-sized companies and the solution is suitable only for companies with a lot of data. For example, you can have a start-up company dealing with a large amount of data, but only have 10 people working on it. In this case, you will need a solution with this capability.

I am satisfied with the support.

If you have used the solution previously then the initial setup of Dell PowerScale (Isilon) is easy, if not then it is difficult. If you have good planning and preparation then the implementation can take two to three days.

We do the implementation of this solution for our customers.

The price of Dell PowerScale (Isilon) is reasonable.

My advice to others is they should work with a provider that knows the solution well and the features in order to implement it correctly.

This is a good solution but it is not always the best choice, it depends on the use case.

I rate Dell PowerScale (Isilon) an eight out of ten.

We have massive unstructured data. It's only used to keep up with our data growth.

Its scalability has been huge for us. Before that, we were using Windows File Servers and there were a lot of labor-intensive log-on activities to build all those servers. With Isilon, we just add nodes and grow capacity. We realized those benefits as soon as we had Isilon online.

We love the scalability with OneFS. It is a one-file system that just grows. It is able to keep up with our massive data and ingestion.

We don't use applications on Isilon, however, it manages the growth of our data and structured data.

The solution's flexibility for supporting various data workloads while keeping them protected is okay. It does the job. I didn't really think about it in terms of protection with its resiliency and its ability to grow.

It's important that PowerScale helps us secure data from cyber attacks, however, budgets control everything we do, so we can only use it as far as what our budget allows.

We lost our technical sales reps about two years ago. We haven't gotten one assigned to us and we'd love to have one.

We would like to see both performance and security improvements which are in all of the releases. We haven't leveraged S3 yet, however, at some point we're going to leverage S3. We're working towards the 9.0 releases. Therefore, we'd like to see some improvements in the protocols. 

I've been using the solution for five or six years.

OneFS seems very stable. I just wish support would get a little better. I realize with COVID 19 it's been hard to keep people.

The scalability is awesome, this is why we have it. 

There are challenges when providing economies of scale for a large cluster, however, it's nothing. It's hard to quantify that because it is just a cluster, but we've been pleased with the scalability overall.

In the past year, it hasn't been as great. They seem to just follow KBS. Prior to that, they were awesome. I'm hoping they get it together.

Negative

We just used Windows File Servers before Isilon. We didn't have any NAS to speak of. We might have used NetApp, however, that's not a scalable solution. We evaluated others and we chose Isilon.

It is very straightforward. It's a little complicated setting up, however, other than that, you just allocate file shares and maintain security patches, and management is easy. It's pretty similar to other systems. It's very straightforward once it's set up.

I deployed it myself. My implementation strategy was to just do it. I have a template that I modify for every deployment and I rack it a certain way so that we can manage the backend cabling, which is a really big deal. I rack it, set it up, and configure it.

We have it in our major data centers, including Seattle, San Francisco, and Phoenix. I do all the maintenance which includes dealing with failed hard drives. There are always errors popping up. I'm just keeping ahead of all the little things that come up, engaging with support, and so forth.

I did not use an integrator, reseller, or consultant for the deployment. I did it myself. 

We have seen an ROI. For example, we don't build Windows File Servers anymore and that eliminates that labor. We have massive data growth and it helps us keep up with adding nodes. However, we have no quantitative numbers for that to share.

The pricing seems reasonable to me. We always evaluate new storage platforms and we've stuck with Isilon.

The costs involved purchasing the hardware licensing and extending maintenance. However, that's with every product.

We haven't really looked into PowerScale's cybersecurity, including its ransomware protection.

We have some Gen 5 and Gen 6 nodes, and we have five clusters. We just purchased the newer Gen 6, A3000, and H700 nodes in our Phoenix data center.

I'd rate the solution nine out of ten. If there was better support I'd give it ten out of ten.

It's an awesome product. I'd advise others to evaluate all the products and just pick the one that's best for them.

Things are in the details and I've always paid attention to that, however, you have got to handle deployment carefully and think ahead to what could go wrong. If you do that, everything will be fine.

The solution provides file-based access for everyone. It's a simplified platform for user-based access to files. It's also very simple to do data center replication for disaster recovery with Isilon.

It's hard to think back to the beginning when we actually got it and how it helped us improve. Of course, it is leaps and bounds over any Windows-based file share that may have existed back in the early 2000s. However, it's really helped engineers manage and maintain it. It's a very simple platform to work with.

The file-based dumping for SQL backups is great. We use that fairly heavily, especially with the flash-based nodes on Isilon. It's been our go-to platform for user-based file access.

The solution's flexibility for supporting various data workloads while keeping them protected is great. We integrate and have Avamar-based backups with Isilon. The protection is great.

Our main goal is to do disaster recovery with whatever solution we use and Isilon makes it pretty simple to replicate those workloads over to our secondary data center.

Cyber security, including ransom protection, is good. We haven't really leveraged a lot of those features as we should. We know that they're there and we work with our partners to help us implement those pieces for us.

The impact PowerScale has had on our organization's storage efficiency is positive. It's tough to talk about efficiency, as we love using it and we dump everything into it. What it retains gives us a great DD compression on the array. We find ourselves overusing it, however, we do have it plugged into Cloud IQ. That helps us with alerts to let us know when we're getting close to our thresholds for capacity.

PowerScale has helped us free up our employees' time to focus on other business priorities. They're not spending a lot of their time managing user files. Everything's managed from the array itself.

PowerScale helped reduce our overall risk. It helped us reduce our overall risk mainly due to the fact that we're replicating between data centers. We don't have to worry about a single point of failure within our data center. I can check on the health of our arrays really at any time with CloudIQ, and everybody sleeps better at night.

If they integrated some functions, as they have on Data Domain with a cyber recovery vault, it would be ideal. They have immutable snaps that they can leverage, however, it would be nice to have something folded in with CyberSense where we could detect points in time when we need to do recovery for anything that may be compromised.

We've had Isilon for many generations. I'd say we've used it for the last eight to ten years.

We have had issues with internal load balancing between some of the shelves. That is an ongoing issue that support is trying to address. We're still waiting on a resolution for that, but that's really been our only issue with the stack that we have.

We haven't had any issues with technical support.

Positive

In terms of what we used before, I can't remember that far back. If we did it was probably some sort of Windows-based file share that was all manual controls. It's hard to compare what we used in the past as we did not switch from another vendor.

I was not directly involved with the initial development of the solution. My involvement was just the management of the storage engineers managing the system. We like to keep everything in the Dell ecosystem so it was easy enough for us to turn it over to the storage engineers.

It wasn't complex at all. Even the upgrades and controllers that we've done have been pretty straightforward, however, we've got two guys today that manage the environment.

It's deployed across two data centers.

The solution does require maintenance. We do continuous controller updates and the like, however, we do it in conjunction with support.

We have seen ROI, however, it's tough to compare since we haven't really looked at other platforms. It's been easier for us to use the platform we have.

The price is reasonable. The ease of use with the array and the functions that we get from it still seems to be better than other products that are on the market today.

I'd rate the solution eight out of ten. It's still great, however, there are some internal issues that need to be resolved. That said, I understand it's always under continuous development.

I would advise potential users to not focus on the price tag right away. I know there are other cheaper solutions, however, they may not have the functionality that Isilon has and the same tie-ins that you'll get with the Dell ecosystem with features like Data Domain and Avamar, et cetera.

We use this solution for the primary file storage for our organization for both the government and health department of Nova Scotia. We use the data domain for backup along with Isilon for our file storage.

This solution offers flexibility in supporting various data workloads and is very easy to work with. It is our go-to solution for all data storage. It is stable, it runs in the background and our users are barely aware that it exists. It operates seamlessly and this is the biggest benefit for our business. 

PowerScale has made it extremely easy to scale file data across our organization. We have two implementations of Isilon. One is a replica of the other. When scaling, we add nodes to each location and expand the cluster. The process is straightforward.

We have eliminated the need for any silos in our data center using this solution. It has also freed up our employees' time to focus on the other business priorities because it is easy to manage and add nodes when necessary. Replication works seamlessly. We are able to meet all of our auditors and governance requirements.

The way it scales easily, is easy to use and its security are the most valuable features. 

It would be nice to see tools like Superna Eyeglass built into PowerScale.

I have been using this solution for ten years. 

This is a stable solution. 

This is an easy solution to scale. 

The solution's service and technical support are excellent. I would rate them a ten out of ten. 

Positive

Many years ago, before Dell, we used a solution for file sharing and office servers. We moved to storage area networks, and the transition to Dell seemed like the right move. 

The initial setup was straightforward. We did receive some assistance from Dell for the setup and they continue to assist us with maintenance on a regular basis.

A reseller assisted us with the implementation. They were excellent. To deploy this solution, we did not require more than our existing operational staff. These three team members take care of storage and backup.

Our biggest ROI in using this solution is the ease and centralization of management. This has freed up team members to focus on other tasks in the data center.

The pricing for this solution is reasonable. 

We evaluated a lot of other alternatives and this solution came out on top. 

I would recommend this solution to others. 

I would rate it a ten out ten. 

We use this solution to facilitate sharing data access across multiple platforms. We are a children's hospital and have a lot of PHI data that is critical to keep secure. 

One of the benefits that we have seen from our research department is quotas and chargeback. They are able to control costs based on the projects that they're given and the grants that they receive from the state and federal levels. They are able to track the quotas and chargebacks, which is made possible through Isilon.

Implementing Isilon has removed the previous silos that existed between different teams. Everyone has been able to virtually separate their resources, but still store them physically on the same box.

PowerScale helped free up our employees' time to focus on other business priorities. There are now automated jobs such as backing up and replicating data, that reduce the footprint we have. Those types of tasks were previously done manually.

Isilon also makes it possible to delete large amounts of data and fix active directory permissions. Previously, we would have to create scripts and run them manually. It also reduced our risk of data loss and gave us the ability to recover from snapshots and replicated data.

We have data that is accessed from multiple OS from different models and in departments in our company. The ability to serve up that data to all those different platforms is very useful.

One of the best features of Isilon is its reliable performance and ability to report on its performance. Reliability is really important in our environment, with a 24/7 shop that serves patients. In many instances, data access is critical.

Prior to Isilon, we had to access data from multiple different platforms. This solution offers unified storage and the ability to consolidate and migrate data which was a big step forward. It allowed us to cut costs by eliminating multiple platforms, putting it all on one array.

Additional metadata reporting would be great. We have to use a separate tool to report on that. We would like to view the age of data and how long it has been since someone has accessed a file.

I have used this solution for eight years. 

This is a stable solution. 

This solution's scalability in an on-premise environment is impressive. We continue to throw large workloads at it and performance has been pretty stable. It has multiple nodes, which is useful when we have outages or code upgrades. We're still able to perform those without interruption of service.

The EMC field support is great. They're easily accessible. We have a specific person we call which is invaluable. We are able to open tickets online instead of spending hours on the phone, no matter what day or time. The only challenge we sometimes experience is a language barrier. 

Positive

The initial setup for this solution is complex. The F900 uses Dell PowerEdge Servers instead of the traditional nodes. We needed to disable memory allocation features on those servers. When we did that, with EMC support, it brought the cluster down and it was down for a couple of weeks. 

The deployment involved a storage analyst, data center analyst, and EMC staff. The data center analyst handled the power requirements and cabling requirements. There are 15,000 users across multiple sites. 

This solution requires three people to handle maintenance. Maintenance requires verifying whether jobs are successful, identifying failures, and ensuring that replication is occurring correctly.  We do regular creation and deletion of shares, files, and folders.

We are able to better handle and reign in budgets by making departments responsible for the data that they are consuming for the grants that they get. The deduplication of data has freed up some of the storage costs that we've traditionally experienced. Some of the newer technology allows us to store more data on less equipment, which means that we're using less footprint in our data center.

This solution is priced slightly higher than others on the market but does offer good quality. With this solution's data reduction and compression, we were able to purchase less. Costs have dropped because of the data rate of compression and deduplication.

We evaluated Pure Storage but their support was unreliable. We need fast and reliable support, and EMC has always proven that when we have an outage, they're there to help us.

The user interface is very simple to use. Support is critical when deploying this solution. When we were deploying the F900, there were a lot of problems that were beyond our scope. We frequently needed to touch base with system engineers from EMC. 

I would rate this solution a nine out of ten.

PowerScale (formerly Isilon) is effectively a giant NAS. We have two clusters, one for production workloads and one for Disaster Recovery and Business Continuity purposes. These clusters are installed in separate data-centers, physically located in two different places in the country. Both clusters were deployed at the same time when we first adopted the solution, and we have been growing them at an almost equal rate ever since.

Our production cluster is attached to our High-Performance Computing (HPC) environment, and this was the primary use case in the beginning: to provide scale-out storage for the Bioinformatics team, who do omics analysis on plant and seafood organisms that we do scientific research on. As time went on, we expanded our use of the platform for other user groups in the organization.

Eventually, PowerScale became the de-facto solution for anything related to unstructured data or file-based storage. Today, we also use the platform to host users’ home directories, large media files, and really any kind of data that doesn't really fit anywhere else, such as in a SharePoint library or a structured database. Nowadays, almost everyone in the organisation is a direct or indirect user of the platform. The bulk of the storage, however, continues to be consumed by our HPC environment, and Bioinformaticians continue to be our largest users. But we also have data scientists, system modellers, chemists, and machine-learning engineers, to name a few. 

Our company has multiple sites throughout the country and overseas, with the two primary data-centers supporting our Head Office and most of the smaller sites. Some of these sites, however, have a need for local storage, so our DR/BCP PowerScale cluster receives replicated data from both our production cluster as well as these other file servers.

Before PowerScale we used to have a different EMC product. I believe it was VNX 5000, which is primarily a block storage array with some NAS functionality. We did not have a HPC environment, however we did have a group of servers that performed approximately the same function.

Back in those days, raw storage had to be partitioned into multiple LUNs, and presented as several independent block devices because of size limitations of the storage array. When one of these devices started to run out of space, it was extremely cumbersome and time-consuming to shift data away from it, which slowed down our science. We wanted a solution that would free our users from the overhead of all of that data wrangling. Isilon was a good fit because it enabled us to effectively consolidate five separate data stores into a single filesystem, providing a single point of entry to our data for all of our users.

PowerScale helped us consolidate our former block storage into a full-fledged, scale-out, file storage platform with great success. We then decided to expand our use cases further, replacing some of the ancillary Windows File Servers that provided network file shares in our Head Office. We now have a single platform for all our unstructured data needs at our main locations.

We have not explored using PowerScale cloud-enabling features yet, but it is in our roadmap. The fact that those features exist out of the box, and can be enabled as required is another reason the platform is so versatile.

The switch to PowerScale was transformative. Before we implemented it, users had to constantly move their data between different storage platforms, which was time consuming and a high barrier of entry for getting the most of our centralized compute. Distributed, parallel processing is challenging enough, to add data wrangling on top of it created massive cognitive overload. Scientists are always under pressure to deliver on time, and deadlines are unforgiving. The easier we can make leveraging technology for them, the better.

We officially launched our current HPC environment shortly after we introduced Isilon, supporting approximately 20 users. Today, that number has grown 17500% to over 350 users across all of our sites. In an organization with nearly 1,000 employees, that's more than a third of our workforce! I credit PowerScale as one of the critical factors responsible for that growth. PowerScale simplified data management because it allows you to present the same data via multiple different protocols (eg: SMB, NFS, FTP, HTTP, etc), tremendously reducing our users’ cognitive overhead.

Before adopting PowerScale, we also faced capacity constraints in our environment. I had to constantly ask end-users to clean up and remove files they no longer needed. Our block data stores were constantly sitting at around 90% utilization. Expanding the storage array was not only expensive: every time that we wanted to provision additional space we had to decide if it was justified to re-architect the environment versus adding yet another data store. And going with the later option meant going back to our users again to free up space before more capacity could be added. All of this wasted massive amounts of time, that could have otherwise been spent running jobs and doing science.

Once we introduced scale-out storage, capacity upgrades and expansion became straightforward. The procurement process was simplified because now we can easily project when we will hit 90% storage utilization, and our users have visibility of how much storage they are individually consuming thanks to accounting-only quotas, which help keeping storage usage down. PowerScale provides a lot of metrics out of the box, which are easy to navigate and visualize using InsightIQ, and most recently DataIQ.

I can certainly recommend PowerScale for mission-critical workloads, it is a powerful but simple platform with little administration overhead. We use it in production for a variety of use cases, and it would be hard for our organization to operate effectively without it.

When we selected Isilon as our preferred storage provider, many considerations came into play, but the deciding factor was how little administration it requires. We no longer need a dedicated storage administrator looking after it. Instead, our Systems Engineers can handle the day-to-day operations without requiring in-depth expertise in storage management. The simplicity of the solution was a strong selling point when we first started looking into it. For example, when you have replicated clusters, you must ensure that you can actually failover between them in the event of a disaster. PowerScale makes setting up and checking the status of replication schedules extremely simple.

Over time, we started using more and more of its capabilities. I believe the most valuable feature we started using, beyond the initial scope for the solution, is the multi-protocol system that allows you to access the same set of files using different network protocols like NFS or SMB. PowerScale’s Unified Permission Model ensures that data security and access permissions are honoured regardless of whether the client is a Windows desktop or a Linux server. Our users can now access the data they need for their research, without having to deal with multiple credentials depending on the environment they are using, or having to rely on specific clients. The same file can be opened and edited from Windows Explorer or from the Linux command line, and we can guarantee that the ownership and permissions of that file will remain consistent. It reduces friction and cognitive overhead, which is what I value the most.

Data security and availability are also included in solution, out-of-the-box. Of course you still need to be aware of how to configure the different features to your use case, but from a data security and availability perspective, you can leverage replication schedules, snapshotting, increased redundancy at rest, and all of those features which we now consider a must-have. With PowerScale, I can have piece of mind that if a specific directory needs to be protected, it will be protected.

The only thing that I think PowerScale could do better is improving the HTTP data access protocol. At the present, you cannot protect access to data via HTTP or HTTPS the same way that you can secure data access through other protocols like NFS or SMB. You can either access a file because it can be access by anyone in the organization, or you cannot at all. There is no in-between. HTTP is not considered a first-class data access protocol, so the Unified Permission Model that would allow a user to authenticate before being able to access a private file, does not apply.

However, with the recent introduction of S3 starting from OneFS 9, I believe the necessary plumbing is already there for HTTPS to also be elevated to a first-class protocol in the future because both protocols sit behind a web server under the hood. It does not sound like it would be too complicated to implement, but it would be a valuable feature and it is currently missing.

We started exploring storage solutions for our environment back in 2012. We have been using PowerScale for nearly 10 years now.

PowerScale has never failed us. Since it was first installed, it has been running with almost 100% uptime since we started using it. We have only had to shut down the entire cluster once because we were moving data-centres. In earlier versions, sometimes you had to reboot the entire cluster for significant OS upgrades. Today, rolling upgrades are the norm, where only a single node is ever down at a time.

At the beginning, we procured four initial nodes, which amounted to about 400 TiB of usable space. We now have just shy of 2 PiB of total installed capacity at each cluster. Our storage usage has grown quite a bit, moving from terabytes to petabytes, but I have no doubt that we will be able to continue growing at the same rate or even more in the future. The original Isilon had already been designed to scale to multiple petabytes, PowerScale will only continue to push that further. We highly value being able to grow our capacity without having to be concerned with platform limits.

PowerScale now also offers more choice when it comes to mixing and matching different types of storage nodes within the same cluster. For example, you can get all-SSD or NVMe nodes alongside old-fashion SAS disks, that you might want to consider adding when performance is critical in your environment. In our case, the performance we get without these new nodes is sufficient for our needs. The best part is that should we ever need to provide a faster pool of disks, there is no administration overhead to do so: just add the new node types, set the tiering rules that you want, and let the system rebalance itself. No partitioning, no moving data around yourself. It is transparent to the end-users as well as the administrators. You can even tier data to a cloud pool for the archive if you want! This simplicity is, again, one of the main reasons we decided to stay on the platform.

I needed technical support on a few occasions, specifically while implementing multi-protocol access for Linux and Windows clients. There was an instance when my engagement with support had to run for longer than I expected, but that was because the solution I wanted to achieve was highly complex from a technical perspective. We had to escalate the issue a few times to the next tier of engineers until they came through with a solution. It was always an excellent customer service experience, and I can certainly recommend Dell EMC Support to anyone who asks.

That said, we only tend to contact Support when we are unable to resolve issues or find the answers with need in the product knowledge bases, or the community forums. The availability of product information online is both comprehensive and of excellent quality.

The initial setup was straightforward. Since it was a green-fields implementation, we did not run into any issues. EMC, who later merged with Dell to form Dell EMC, even let us evaluate the platform in our own data-centre, so by the time we decided to procure the solution, all we had to do was to revert to “factory settings”. The longest part of the process was migrating around 84 TiB of data from our old data stores, as it happens with any data migration exercises. But once the data had been relocated, it became a matter of simply pointing the servers to the new data store entry points. Users were happy to take it from there, and were certainly overjoyed at the additional space they had to work with.

It was a long time ago now so details are fuzzy, but we dealt with EMC directly, with the help of an integrator for some of the initial design and implementation. EMC was our primary point of contact for platform-specific support when we first started, and their guidance around the different features of the platform was invaluable.

Today, that same integrator continues to help us with ongoing procurement, simplifying decisions around which of the many available node types might be the best suited to our environment, or ensuring that we stay on top of our node refresh cycle as older ones reach end of life.

Price was also a significant factor in our decision to go with PowerScale. The team at EMC, now Dell EMC, came through with a highly competitive offer that tipped the scales towards their solution. There was only one other solution around the same price point, but it could not match PowerScale on features. That other solution is no longer on the market.

The licensing model is interesting, because it is essentially “pay to unlock”. Most of the available features are software-defined, so they are already available in OneFS, the underlying Operating System, waiting for you to activate them as needed. There are a few additional costs, however. NDMP backups require you to install fibre cards, which are sold separately. Then of course you have the cost of tape and off-site storage, but you would have those same costs with most other platforms. Luckily, we do not need to back-up the whole cluster because we can rely on cluster replication and snapshots (on both source and target clusters) to achieve our RPOs. But we do have a legal requirement to preserve some data for an extended period, so we use tape for that.

We evaluated three other competing solutions based on multiple criteria. Some of those solutions no longer exist, or have evolved into a different offering. We went through a rigorous evaluation process, which assessed the platforms’ scalability, ease of use or complexity to administer, performance, and of course TCO. Isilon was the brand name that blew all others out of the water. It was an easy decision for us to make based on the criteria we set.

I give Dell EMC PowerScale a high 9 out of 10. It is not quite a 10, mainly because we do not have a use for all the features it provides, which you need to be aware of from a security point of view (eg: to ensure that they do not introduce unexpected risk). The ecosystem has also grown to be somewhat more complex in terms of the many different types of nodes that you can have. This gives you a lot of flexibility, but it does go slightly against the idea of simplicity that was so attractive initially.