Dell PowerScale (Isilon) Reviews

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 use the solution to organize the data structure. Some of its applications are geared towards companies in the oil and gas sector. For instance, it supports SIP solutions that conduct scanning and comprehensive Seismographic analysis. Additionally, other customers include broadcast companies with vast historical assets. Essentially, they aim to manage their content libraries efficiently. It primarily focuses on data management and storage solutions.

PowerScale addresses the customer's need for a global rather than discrete file system. It resolves performance issues and offers comprehensive support. PowerScale needs more expansion regarding solutions such as HSM or integration with tape libraries.

Dell has pairing and utilizes optical services within the same infrastructure. This means utilizing services from the same infrastructure for internal file system needs and providing access to the public.

The solution should improve its pricing and features.

I have been using Dell PowerScale (Isilon) as a consultant and reseller for seven years.

The product is stable.

The solution is scalable and is suitable for enterprise customers.

Support is very good.

Positive

IBM is cheaper than Dell PowerScale.

The maintenance depends on the time you are willing to invest in learning about the platform. It varies for each individual, and if you have people eager to learn, it can make a significant difference.

IBM built its sources of disk management which control costs. They don't rely on purchasing from vendors. For example, Dell PowerScale doesn't manufacture the disks; instead, they source them from suppliers or engage in patching. They do not produce the disks themselves; they procure them.

IBM can utilize gateways that offer a similar file system to PowerScale. These gateways provide both block storage and file services. This is different from PowerScale because when purchasing PowerScale, you acquire building blocks including CPU and memory. This configuration lacks the flexibility to adapt to various infrastructures. While this setup can be configured, it may pose limitations.

You can customize security settings within the tool, including access and file-level permissions. This focuses on enabling 'write once' capabilities, making it challenging to alter data without appropriate authorization. It would be impossible to tamper with unless an individual gains access by obtaining administrator credentials.

Overall, I rate the solution an eight out of ten.

• Research data
• Departmental file shares
• Data centre storage: NFS

We have two data centres in our university. We have Cisco UCS, Pure Storage, and are heavily virtualised with VMware. PowerScale is our unstructured data storage platform. It provides scaled-out storage and our high-level NFS across applications. It also provides all the storage for our researchers and business areas, as well as students, on the network.

With the exception of block workloads, which is primarily VMware, Oracle Databases, etc., everything else it is on PowerScale. It definitely has allowed us to consolidate the ease of management.

With the quotas having fewer large pools of storage in the data centres, we typically only have one or two Isilon clusters. That gives us the ability to multi-tenant or allocate data to different applications and isolate workloads. It is very efficient when managing that volume of storage. We are not tuning it every day or week. The only time that we are really doing anything with it is if we're planning an upgrade of some sort several times a year. Outside of that, it just does what we want it to do. 

We automate the vast majority of the things that we do on the Isilon clusters: provisioning of storage, allocation of storage, management of quotas wrapped into tens of thousands of students, and managing permissions. That's the level of support they have for their built-in API's, which is probably a huge game changer for us in the way that we manage the storage. It makes it far more efficient inside of PowerScale.

Compared to doing it manually, what we have been able to automate using the API is saving us at least tens of hours a month versus when we used to get service requests. We have even been able to delegate out to different areas. If we have an area with whom we do file shares, we delegate out the ability for them to create new shares and manage their permissions themselves. 

The solution allows us to manage storage without managing RAID groups or migrating volumes between controllers. We see this in the big refresh that we did earlier in the year. After you have clicked the "Join" button and joined, you go to the old node and click remove, then wait for it to finish. You don't have to configure anything when you add new node types, they are automatically configured. You can tune them and override things if you want, but there is no configuration required.

PowerScale has enabled us to maximise the business value of our data and gain new insights from it. It gives us the ability to have our data stored and presented via whatever protocol is required. Now, we can look at all these different protocols without having to move or duplicate the data.

The solution allows you to focus on data management, rather than storage management, so you can get the most out of your data. We looked at the types of data that we have on the cluster, then we just target it based on the requirements. We don't have to worry about building up different capabilities, arrays, RAID types, etc. We just have the nodes, and through simple policy, can manage it as data rather than managing it as different RAID pools and capacity levels. If someone needs some data storage, then we ask what their requirements are and we just target based on that. Therefore, we manage it as a workload rather than a disk type. 

Their SmartQuotas feature is probably the thing that we use most heavily and consistently. Because it is a scaled-out NAS product, you end up with clusters of multiple petabytes. This allows you to have quotas for people and present smaller chunks of storage to different users and applications, managing oversubscription very easily.

We use the policy-based file placement, so we have multiple pools of storage. We use the cold space file placement to place, e.g., less-frequently accessed or replicated data onto archive nodes and more high-performance research data onto our high-performance nodes. It is very easy to use and very straightforward.

The node pools give us the ability to non-disruptively replace the whole cluster. With our most recent Gen6 upgrade, we moved from the Gen5 nodes to the Gen6 nodes. In January this year, we ended up doing a full replacement of every component in the system. That included storage nodes, switching, etc., which we were able to replace non-disruptively and without any outages to our end users or applications.

We use the InsightIQ product, which they are now deprecating and moving into CloudIQ. The InsightIQ product has been very good. You can break down the cost performance right down to protocol latency by workstation. When we infrequently do have issues, we use it to track down those issues. It also has a very good file system reporting.

For maximising storage utilisation, it is very good. As you add more nodes in a cluster, you typically get more effective utilisation. It is incredibly flexible in that you can select different protection levels for different files, not necessarily for file systems or blocks of storage, but actually on a per file basis. Occasionally, if we have some data that is not important, we might need to use a lower protection. For other data that is important, we can increase that. However, we have been very happy with the utilisation.

Dell EMC keeps adding more features to the solution’s OneFS operating system. In terms of group work, we have used it for about 13 years. The core feature set rollup has largely stayed the same over that time. It has been greatly improved over that time as well. So, it has always been that storage NFS sandbox, and they've broadened their scope for NFS v4, SMB3 Multi-channel, etc. They are always bringing up newer protocols, such as S3. Typically, those new features, such as S3, don't require new licensing. They are just included, which is nice.

Over the years, the improvements to existing protocols have been important to us. When we first started using it, they were running open source sandbox for their SMB implementation under the covers and they used a built-in NFS server in a free VSD. Whereas, with the new implementations that they introduced for OneFS 7 have had huge increases in performance and been very good, though there's not necessarily any new features. We even use HDFS on the Isilons as well at the moment. The continued improvement has been really beneficial.

It is incredibly easy to use the solution for deploying and managing storage at the petabyte scale. With CIFS and IBM Spectrum Scale, there just isn't the horizontal concern. I couldn't think of an easier way to deploy Petabyte NAS storage than using Dell EMC PowerScale.

The replication could lend itself to some improvement around encryption in transit and managing the racing of large volumes of data. The process of file over and file back can be tedious. Hopefully, you never end up going into a DR. If you do go into a DR, you know the data is there on the remote site. However, in terms of the process of setting up the replicates and filing them back, that is just very tedious and could definitely do with some improvement. 

There is a lack of object support, which they have only just rectified. 

About seven years.

The stability has been exceptional. I've been very happy with the stability of it. In the last six years, we have pretty much been disruption free. Prior to that, we have had one or two issues, which we worked with their support to fix. 

We had a major refresh at the start of the year when we replaced one petabyte at one site and a half a petabyte at another site. This completely replaced everything and took us about a month. It was finished with one staff member overseeing the process, moving the data and roping in one or two other staff at different times to help with the physical backing. 

They are quite heavy, so you always want to have two or three people involved. It has very minimal staff management required. For example, once the hardware is racked, it needs just one operator who joins the nodes, waiting for the data to move over. Internally, this is non-disruptive to the user. 

Firing up the old nodes, that is more of a management thing. 

Pretty much everyone touches the solution in some way or another. It has been a bit different right now with COVID-19, since a lot of people have been recently working remotely. In any given day, probably 12,000 people have been using it. That is just going by the number of active connections that we have from staff, students, and researchers at any time.

We can't see anyway that we would ever reach the limits of the product in terms of scalability and our workloads. We have no concerns around scalability. 

It has a back-end network that it's managing to get switches with enough ports to plug the nodes in, if you want to go big. That is the most complicated part, not the actual management of storage. As you add more nodes, that management overhead remains largely the same. 

For larger scalability, I would be very comfortable with it. We would just have to do some good site planning to ensure that we have enough room for it.

Our usage is pretty extensive. It touches on almost every area of our organization. With the introduction object and support for Red Hat OpenShift, which they're releasing in OneFS 9.0, we are very keen to explore and extend the usage in those areas. That is part of the reason why we are upgrading our test cluster on OneFS 9.0 to specifically evaluate use with Red Hat OpenShift and Kubernetes in clouds. It definitely has a very strong place now in the data centre, and we don't see it going away anytime soon, as we see more workloads going onto it.

The support has been mixed. If you get through to the right engineers, you can get problems resolved incredibly quickly. If you don't, you can go around in circles for a long time. We do typically have to escalate support tickets through account managers to get them positioned correctly. However, once that happens, issues are resolved pretty quickly and we're generally happy. 

The technical support is average. There are certainly not the best that we have ever dealt with, but far from the worst ones. I would not recommend the product based on their tech support alone. 

Going back 13 years prior, we used to have a lot of Microsoft and Linux-based file servers all over the place. They were all siloed with a lot of wasted capacity. Consolidating all those down into a small handful of Isilon clusters has dramatically reduced the amount of silos that we have in the organization. In terms of reducing waste from having storage stuck in one silo or isolated area, it has made a huge improvement.

We have previously used IBM Spectrum, and I don't think you can buy anymore. Briefly, eight years ago, we moved a large portion of the workload off Isilon onto Spectrum. That was the biggest regret that I have had in my career. We couldn't get back on the Isilon fast enough. It was a commercial decision to move away from Isilon, which wasn't the cheapest. However, it was far more mature than the IBM product. Spectrum cost us so much that what we saved in capital expenditure we then lost in productivity, overhead, and maintenance. It was just a disaster. The support that we received from IBM was the worst support I have ever received. I've been in this industry and job for about 17 years now, and I have never had a worst support experience that I've had from IBM. It was a nightmare.

When we needed to get the issue with Spectrum fixed, there was no doubt about getting PowerScale. We couldn't get back on PowerScale fast enough. We just made that happen, and as soon as we did, all the fires were put out.

About 13 years ago, we were using six terabyte nodes back. Now, they're obviously a lot bigger than that. While scalability was definitely a key interest, the main driver for us was the ease of management to sort of consolidate all the separate file servers with their own operating systems and RAID arrays, and consolidating them into one pool of storage where we could allocate quotas and still manage capacity effectively, but centralize it and reduce waste. The ability to scale out was just icing on the cake, and definitely something we were very interested in. It's something we've utilised quite heavily over time, but the ease of management was the main driver.

The initial setup has always been straightforward. The process of creating a new cluster is largely the same now as it was 13 years ago. You get your first node, then connect the serial port to it. You answer about 10 questions, then you're ready to go. The rest of the nodes are added by clicking a button. It's incredibly easy to set up, and it says a lot that the process has been the same for about 13 years. There's not really much to improve or simplify, because it is already incredibly simple.

Assuming the hardware was racked, you could have the cluster setup and your minimum three nodes joined within half an hour to 45 minutes.

The process of adding a node is very straightforward: It is pressing a button. This can take five minutes, then the process is complete. Once you have added new nodes, you can then remove old nodes. 

Understand your workload. Make sure you size and cost it correctly for the amount of metadata you expect to see on it. Don't undersize your SSD.

For the whole replacement this year, I got one of our junior staff members, who had have never actually used our PowerScale, to do the whole upgrade process. I just pointed him in the right direction. Because it was very easy, he managed to do it without any issues.

We don't use any professional services. We always do it in-house. 

Two people are needed for racking hardware. Only one person is needed to deploy it, as that process is very straightforward.

The solution has simplified management by consolidating our workloads. Rather than managing all the different workloads on different storage arrays, Windows Servers, etc., we just have one place per data centre where we manage all their unstructured data, saving us time.

PowerScale has reduced the number of admins that we need. It has allowed our admins to focus on adding value through automating tasks and streamlining operations for our customers, rather than focusing on the day-to-day and tuning RAID profiles. We can use our APIs to automate workflows for customers and have quicker turnaround times.

The solution is expensive; it is not the cheapest solution out there. If you look at it from a total cost of ownership perspective, then it is a very compelling solution. However, if you're looking at just dollar per terabyte and not looking at the big picture, then you could be distracted by the price. It is not an amazing price, but it's pretty good. It is also very good when you consider the total cost of ownership and ease of management.

We added on a deduplication license. That is the only thing that we have added. That was a decision where it was cheaper for us to license the deduplication than it was to buy more storage, so we went with that approach. We just did an analysis and found this was the case.

We haven't really hit a workload or situation that we have had any issues catering for. Certainly with the huge number of different node types now, we could position any sort of performance from very cheap, deep archive through to high performance, random workloads. I feel like we could respond very quickly to any business requirement that came up assuming they had budget. Even if we didn't have budget, largely with the way our clusters are configured, we typically mix in high and low performance. We won't buy top of the line, high performance, but we will buy basic H500 nodes, which are a large amount of self-spinning disks. That is what we standardize for our high performance tier. 

13 years ago, it was called Isilon Systems. They were a start up in Seattle, while we are in Australia. We were importing the hardware directly. At that time, there was nothing really else that we were looking at. We were just caught up in revolutionising the way we would be managing one pool storage. Then, six to eight year ago, when we had that little stint on IBM Spectrum, we didn't go to market. We very heavily evaluated the IBM product and NetApp in cluster mode as an alternative. We did rule out NetApp from a management perspective as far too difficult to manage. The Spectrum product that we saw on paper and from our evaluation of loaned hardware seemed like it was going to be on par with Isilon. Little did we know the nightmare that would ensue from that. 

The biggest lesson that we learned was from moving away from it onto the IBM product. The maturity of a product is very directly correlated to the amount of time you spend managing it, as it is a very mature product. We have been using it for 13 years, and the core has a very solid, mature foundation that has been built over that time.

We have dealt with Nimble Storage in the past. I would recommend Nimble Storage based on their support (at that time), as they had exceptional support. However, Dell EMC support is no worse than Cisco or any of the other vendors that we have had to deal with, but it is nothing special.

Just don't underestimate how important a mature product is compared to something leading edge or new.

PowerScale's positioned primarily to receive the call within that data centre. We have PowerScale heavily centralized, both in our IT department and on our campuses. We don't really have any storage from PowerScale in the cloud or our edge because we have very good network connectivity. In terms of the right tiers of storage, the level of flexibility that we have for adding different types of storage with different characteristics to our existing cluster now is the best it's ever been in the 13 years that we've managed it. 

Between CloudIQ and DataIQ, they're replacing their legacy InsightIQ product. We haven't moved to CloudIQ yet to start looking at it.

Early on, since we have been using the solution for 13 years, if you added a new node type, then you would have to add three physical nodes to start a new pool and only end up with 66 percent utilisation on that storage pool. Whereas, in the Gen6 hardware, you can have more smaller nodes in one rackmount chassis. Now, you can add a new storage type and gain much better storage efficiency off the bat.

The S3 protocol specifically comes in OneFS 9.0. We have a test cluster for it, which we are in the process of upgrading to have a look at their S3 support. However, I haven't used it yet. Typically, we use something like MinIO, which is an open source object gateway, and put that in front of the PowerScale cluster.

On the archive side, we still have the A200 nodes. While you can go with the A2000s or go deeper than that, we can manage pretty much anything thrown our way by not going too extreme in our pools by positioning data effectively. I think it's very good.

I would rate the solution as a nine out of 10.

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.

We use the solution for NFS. 

Dell PowerScale's performance is good. 

The product needs to improve CLI since commands are complex. The search option is also difficult since you must give the full path. 

I have been using the product for more than three years. 

Dell PowerScale is stable. 

My company has more than 1000 users for the solution, and it is scalable. 

Dell PowerScale's support is good. 

Positive

Dell PowerScale's deployment is complex. 

I rate Dell PowerScale a nine out of ten. 

I rely on Dell PowerScale to manage and store manufacturing data from NES systems. It is great for handling big data, offers scalability, and ensures data security, simplifying the storage and retrieval of manufacturing information.

The most valuable features of Dell PowerScale for its reliable performance and stability. The guaranteed performance, combined with the scalability through its scale-out capability, makes it an excellent choice. It is a top pick for my customers due to these qualities.

There is room for improvement in its handling of object storage. While it excels in managing file systems, enhancing features for more efficient handling of objects could make it even better, ensuring faster and smoother operations.

I have been working with Dell PowerScale for ten years.

The technical support is very good. They are quite helpful.

The installation is quite simple and the deployment takes only a few hours. In my organization, we have a total of 26 system engineers, and among them, 11 are trained to handle Dell PowerScale. These engineers are responsible for installing, maintaining, and providing support for the product. We conduct maintenance regularly, typically with a focus on remote support, offering 24/7 assistance. For on-site support, our goal is to respond within four hours.

While customers often perceive it as expensive, I find that considering its functions and performance, Dell PowerScale is reasonably priced. I would rate it as a five out of ten in terms of costliness. In addition to the standard licensing fee, there are extra costs for services and additional solutions with Dell PowerScale. Services like maintenance and support may incur charges. Sub-solutions such as CloudIQ are also additional and may be charged based on usage or quota.

I would highly recommend Dell PowerScale to others. Overall, I would rate it as a nine out of ten.

We use this solution to manage large data sets. It makes it easy for us to consolidate data storage and multiple applications into a single platform for easier manageability. This means we only need to manage a minimal number of clusters. 

The solution is very flexible in supporting various data workloads while keeping them protected. It scales well with different performance node types. In terms of data protection, it replicates easily with the replication features of SyncIQ. 

We have multiple clusters across two data centers and have two main teams; our enterprise storage services team, and our HPC team who recently purchased a bunch of capacity from Iceland.

When it comes to PowerScale Cybersecurity and ransomware protection, we use a third-party solution called Superna. It is still important to us that PowerScale helps secure our data from cyber attacks. An attack can happen very quickly and be very damaging. Large data sets are difficult to protect, replicate and recover.  

Using this solution, provisioning is a lot easier because we have large capacity clusters and can easily provide storage space to users as needed. We realized the benefits of this solution shortly after deployment. The management of the solution is very easy. 

This solution has had a positive effect on our company's storage efficiency. We are able to add capacity as needed in order to scale.

The ease of use of this solution has been the most valuable aspect as well as the SyncIQ and snapshot features. 

It assists with eliminating storage silos because it provides SMB and NFS protocols. PowerScale has also helped free up our employee's time to focus on other business priorities. 

I'd like to see more Iceland products in the cloud so that we can port our data into different environments if needed. I would also like to see a virtual appliance or software-defined Iceland product.

Version upgrades and patches take a long time to complete. This could be improved. 

I have been using this solution for twelve years. 

This is a stable solution.

This is a scalable solution from a file data perspective. Scalability can be more difficult because you need three nodes minimum to start a cluster. You need a lot of other hardware to provide that service. This solution is not scalable in the cloud.

 I would rate the support for this solution a seven out of ten. The quality of support depends on the agent you deal with. 

Neutral

We have used other solutions but Iceland is the storage platform of choice for the large data sets that we have. We have a very strong relationship with Dell and their support is very good so we were very comfortable choosing Iceland.

The initial setup is straightforward. Dell came in to install the solution and update the operating systems. After that, some administration and management tasks needed to be completed, and then the setup was complete. 

Pricing for this solution is reasonable. 

This solution scales very easily. You just need to be aware of how much you scale as this can complicate the management of the solution. Scaling requires a strategy. 

PowerScale is easy to manage but more difficult to maintain. This is because it is a multiple node environment and the larger you get, the longer it takes. There are more risks when you make a change.

I would rate this solution a nine out of ten. 

We used it originally for archiving our video storage, and then we expanded it to include user shares. All of our unstructured data has been moved to PowerScale.

We have now expanded the OneFS to start to use Local S3 Buckets, that use the same API setup as Amazon, but lets us host the data onsite.

In addition we added Power Protect Data Manager that is allowed to backup the Shares and stores, allowing us to have a backup of everything on another location.

We moved our shares over. Now, instead of taking up a large amount of space on a virtual machine, our shares take it up on one appliance. The load on that virtual machine is much less and it makes it easy to future-proof it, because now we don't have to move it again in the next migration of servers.

We have saved about 30 percent on storage with it. And as we grow, we get more space, meaning the efficiency improves each time we add a node. We went from 75 percent efficiency to 82.5 percent efficiency when we expanded.

The solution provides us with the flexibility to add the right tier of storage at the right time for data that resides at the edge, core, or cloud. That really is nice. We did one use case where we put it out at the edge, and it was nice to have the Isilon at the edge. It really helped improve things. It helped the storage of the cameras, and it helped get the data back to the core in a reasonable time. It allowed us to go from the edge to the core and then up to the cloud, instead of trying to go from the very edge to cloud.

PowerScale also allows us to manage storage without managing RAID groups or migrating volumes between controllers. It simplifies the storage. It allows us to see everything as one large volume instead of having multiple volumes all over the place.

And when it comes to the business value of our data, it allows us to see what's being used and how it's being used, and we can do so much more quickly and efficiently. As a result, we can better evaluate how we're storing the data.

It has also helped us to reduce data silos. We used to have four video servers out there, all storing data. On the home farm, now, we're down to one server storing data in one location, and that includes all the user shares. 

All our data is in one place and that has increased performance. We could never afford to say, "Let's have this information on solid-state," and allowed the OneFS to decide, based on usage, of where it would be stored: on a fast drive or on a slow drive. It automatically does that in the background for us, instead of our having to manually move it and then have the user change where they get the information from.

In addition, it has simplified management by consolidating our workloads. It's all done in the same portal now. And while it hasn't reduced our number of storage admins, it has definitely reduced the time we spend looking at it, so we can focus on other efforts. It saves me about five hours a week.

Another benefit is that it allows us to focus on the data rather than where it's stored. Now, we don't have to worry about moving it around from place to place to get efficiencies out of the data. We just have it all in one place. The single interface, the SmartPools policy, decides where it needs to reside.

The single pane of glass for both IT and for the end-user is a valuable feature. On the IT side, I can actually control where things are stored, whether something is stored on solid-state drives or spinning drives, as well as the access users get. But the end-user doesn't distinguish the difference between a file and its folder; the end-user doesn't have to see the difference.

The single pane of glass makes it very easy to use and very easy to understand. We started at 100 terabytes and we moved to 250 and it still feels like the exact same system and we're able to move data as needed. There are no performance issues based on how large the storage is.

Adding a node is as simple as racking and stacking the items. It takes about two to three hours to put it into the rack. Once you have it all wired up, it takes you about an hour or 90 minutes with Dell, just to configure things and make sure it's all working. Then you just redefine your policy for where you want the items stored. We just expanded to include the solid-state, a full F200 node, and we just redefined where we wanted those files stored, whether on the super-fast solid-state or on the slow archival mode. Then, overnight, it ran that script and moved all the files around to help increase performance.

We also use the CloudIQ feature to monitor performance and other data remotely. It gives us better insight into where the data's stored and the access times involved. It gives me a better understanding of what's really being accessed and helps me decide what I can move to slower drives first, and what needs to stay in the front-end and remain very fast.

There aren't many templates still coming out for it. They need to provide templates so we can copy and paste what we've done in the past to future, new things.

The refresh of the interface with version 9.3 did help a lot of the things. They are at least improving it.

I have been using Dell EMC PowerScale for about a year and a half.

It's very stable. It's one of the first solutions that I feel comfortable working with during the business day, while people are using it, knowing that I can change things and it's not going to take the system down.

One of the things I like the most about it is the fact that we can scale out now. If we need more space, we order more nodes and it just changes the file structure; it just expands. There are no more individual drives, new arrays, moving things around. It'll just be there.

The future-proofing of what we're doing is a great thing too, because in five years when we're ready to replace that node, just due to its age, we can put the new one in and tell it to archive the old unit. It will move all the files over, in the background, and then we will just remove the old unit. There's no more having to tell users that, "Oh, this whole share is moving and all this stuff is getting done."

The technical support has been really good. It's pretty intuitive to put a ticket in, both through their email and through the calling system. It's usually pretty seamless to get to talk to somebody to actually resolve the issue.

Before PowerScale it was just MD Storage Arrays, the standard, and the LUNs that you'd have anywhere. We eliminated that with this. We originally started with PowerScale for our video system. We were looking for a better system, in the long-term, to store our archival video and process it. We looked at unstructured data solutions and picked PowerScale for that and for the future-proofing.

Also, because we are a large Dell EMC shop, it allowed us to keep it all on the same platform. In looking to do things on a larger scale, it allowed us future compatibility, much more easily. Its ability to meet unpredictable future storage needs looks great. It feels like a great solution and it was the right direction for us.

The first setup was pretty complex and a little different to do. Once we had the core system set up, the next deployment was much easier. The complexity came from changing our thought process, internally, regarding how we store files and how unstructured data really works, and then, how to efficiently use this.

Our deployment took about a week. We did a slow move-over, and we still continue to move anything we find over to it.

In terms of administration of the solution, for the most part it's just me who does a lot of the core work. All the users on the farm are using the system now, meaning about 350 people are accessing the data on the Isilon.

We used the reseller, Dell EMC, for the deployment, and it was a great experience. They were there to help us and make sure we understood where we were going and what we were doing.

The fact that, with PowerScale, we could start with a few nodes and scale very large made it very cost-efficient for us. It allowed us to start out, see what it can do, and evaluate the product before we actually did a larger investment in it. We invested into it again three months later.

I'd like to say we have seen ROI because we're feeling like we're really starting to store data better and understand what's going on, more than we did a year-and-a-half ago.

It's one of those situations where you have to find the right price for you. When we talked to the reseller, we were able to negotiate the right price for what we needed.

We looked at HPE and IBM.

I liked the interface of the PowerScale much better than the other ones. It was more intuitive. I logged on and could almost get to work with it right away. I felt like I could hop on and just start using it, whereas with the other ones I felt that there was a larger, steeper learning curve.

Dell EMC keeps adding more features to the solution's OneFS operating system. The last addition was its CloudPools and that allows us to do backups to the public cloud for the data that we want to keep but don't even need on-prem anymore. It turned the system into a never-ending resource. We can now decide what we want to keep, long-term, without having to expand our storage system.

PowerScale is one of those things that will grow in your environment. Once you start it with one thing, you'll learn that it can do much more, very quickly. That's a great thing about starting small with it, you can expand very quickly later on.