What needs improvement with Dell EMC PowerScale (Isilon)?

The analytics could be improved.

Dell PowerScale (Isilon) is a little bit pricey, and its pricing could be improved.

The price always has room for improvement.

As the product is used for basic purposes, there are not many areas in the tool that require improvement. The product is just fine for now. My company does not use many of the features of the product. I can tell what my customers want to improve in the product, and I believe that they don't want anything to be changed in the solution. The support offered by the product is an area of concern where improvements are required. My customers don't receive much support. The support has slowed down a bit. For Vietnam, the support has to be worked on since it is slow in our country.

Dell PowerScale (Isilon) is a bit expensive compared to other products.

The disk failure handling could be better. We had a few issues because one disk failed, which slightly impacted the whole node, but overall, the performance is stable. We had to expand and increase the storage capacity, but it turned out we needed a totally new cluster because we couldn't expand the current one due to software incompatibilities. We think the backend network should be upgraded from InfiniBand to Ethernet networking, and we had to buy a new cluster. Dell has significantly increased the prices of new nodes, by a lot. As a result, we're considering new vendors because if prices increase by 300%, then the clients will look elsewhere. The disc server handling is not the best. Maybe the tiering logic could be improved. It's based on our age or some other logic but not on disk usage. If you have a tier one, you don’t want to load it completely; it deserves some free space for new data, and you can’t manage that with the current tiering logic. That’s where this software or this solution could be improved.

The solution should improve its pricing and features.

The solution’s interface and pricing could be improved.

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.

The product’s expansion capacity, pricing clarity, and ease of use need improvement. There could be power sizing tools available similar to OneFS or H-Series.

The solution's rate structure or rate redundancy needs to be improved. If you have a lot of nodes, for example, 15 nodes, and when you say you lost one node on the paper, the performance is not affected. However, if you have so much file count and fifteen nodes, losing one node really affects the performance. One of our customers had this issue. Also, when you lose a couple of drives, it's a different structure, then you can lose data. It is clearer on the NetApp side, and you can create a rate group and a pool. So you will have more redundancy on the drives. For example, in a rate set for around 100 drives, you can lose around 20 drives, which depends on the configuration. In NetApp, we will not lose data, but if you lose 20 drives in EMC, you will also lose data. Small systems like Dell PowerScale (Isilon) work perfectly with five, six, or ten nodes. Based on my experience, if you have 15 nodes in Dell PowerScale (Isilon) and lose one node, it creates some problems.

The solution's configuration needs to be more straightforward. Also, its performance could be better than Panasas. In addition, its price needs improvement as well.

The cost of Dell PowerScale is currently high and there is room for improvement.

The price of the solution could be reduced.

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.

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'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.

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.

The management and monitoring tools comprise a disparate suite of products and the roadmap is very unclear. We've got four different products that look after the Isilon, management-wise, and it's a bit of a mess.

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.

The only problem with the WORM (write once, read many) feature is it does take up more space than if you just wrote a file, because it writes stuff twice. But it works for us for chain-of-custody scenarios, and it's built into the file system itself. Also, on the PowerScale system, because of the magic that it does "under the hood," it is very difficult to find out within the system where all your storage is going. That's a little bit of a ding that we have on it. It does so much magic in order to protect itself from drive failures or multiple drive failures, that it automatically handles the provisioning and storage of your data. But by doing that, finding out why a file of a certain size, or a directory of a certain size, is using more storage than is being reported in InsightIQ, is very difficult to discern. It's the secret sauce of protecting your data and that makes it a little disconcerting for someone who is used to seeing if a directory is using 5 MB of space. So if you have a directory using a terabyte of space, it might be using a little bit more because of the way that the system handles data protection. That is something you have to get used to. Also, a lot of people are not used to the tagging or the description in the InsightIQ application. We're used to using the normal nomenclature of terabyte, petabyte, etc. They utilize TB byte and PB byte. So you have to understand the difference when InsightIQ is telling you how much storage you have. It's different than what we're used to. It uses base-2 and the world is used to base-10. Discerning how much storage you actually have, from the information in InsightIQ, takes a little bit of math, but it's not very difficult. I wish they had an interface in there where you could click and it would report in the way the industry is used to, which is in terabytes and petabytes. It's nothing major, just something you have to get used to when you're looking at it.

Some improvements to the NFS support would be of interest to us. I think that will be available next year.

There is room for improvement with the updates. It can take a significant amount of time to do a major OS update. However, even though it takes multiple reboots, the cluster stays up. If we want to apply a newer version of the OS, we have to roll back some of the patches so that we can upgrade. It requires a few reboots just to do that. The cluster doesn't come down, everything is still running, but it's time-consuming, at times.

Simplify where you can. If you have a need for tiering, then that can be okay, but it can behave in ways that you may not expect. If it's at all possible to simplify and stick with one node type, your consistency will definitely stand up a little better. If you do have a workload where tiering makes sense, PowerScale does do a good job of that. That's the only real, "Gotcha," that we've run into.

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.

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 did help a lot of the things. They are at least improving it.

The solution lacks a cloud version. It would be useful if the solution could direct to AWS or Google Cloud effectively or have an AWS version. With the global lockdown conditions, you can't get to the site. It would be easier if it was connected to the cloud. The solution can be a bit complex for those not well versed in the technology.

The solution isn't suitable for small environments or small customers. The price point would be too high for companies that don't need a very large amount of storage space. The redundancy is a little bit expensive.