Software-Defined Storage Products: IT Pro Perspective

“Automatic rebalancing is the feature saving administration time, and also auto rebuilding. Enterprise storage IOPS limiting is also good. The primary goal was to have a storage solution where you do not have to invest a large amount of money just to start, and have idle resources, yet have a locally supported enterprise solution on commodity hardware. In other words, have a pay-as-you-grow, horizontally scaled solution, where front-end controllers are not the bottleneck.”

“There is no built-in system for viewing history data, such as volume IOPS. We have to provide graphing by Prometheus and Grafana, which would be a good new feature in ScaleIO. The next good new feature would be moving volumes between different storage pools, e.g., from a SAS pool to a SSD pool. It would be nice to set minimum IOPS per volume, besides just the maximum, to be able to satisfy this demand from customers out of the box, not by calculating number of disks, etc. It would be nice to have better integration with monitoring and other vendor provisioning and orchestration tools. I am aware that this is a hard-to-achieve task, where it is necessary for product not to be proprietary and to become industry standard.”

“The ScaleIO UI has been working with storage for a long time. Therefore, they know how to clearly present any important data, including data flow and each drive's IOPS/bandwidth, and allow the user to easily monitor bottlenecks and problems, especially the rebuild and rebalance status of child objects. It controls them, as well as maintaining them well.”

“StoreVirtual has been real good for us. We started with the original P4300 LeftHand SANs before they became StoreVirtual. What I love about those is the two nodes and the mirroring back and forth, and you can't lose anything. It's very solid, and we haven't really had any trouble with those either. We have a newer StoreVirtual that we've connected to one of the C3000 Blade Enclosures and it runs well. We lost a system board once and we lost a couple of servers, but we were able to bring everything back. Equipment-wise, it allows us to do all our work. We're real happy with that.”

“The user interface needs to be updated. It's getting kind of long in the tooth, and the user interface makes it look a lot more complex than it actually is to manage, and I think that you can mask a lot of that with a refresh of the user interface. While HPE has created a new HTML5 UI for the HyperConverged 380, it is not available to the rest of the StoreVirtual population.”

“The product helps us to manage our storage in a way that allows us to put different frames inside or out of our storage infrastructure and migrate. The benefits are that it speeds up provisioning of the storage across different tiers and allows a small team to manage that function, for many petabytes of data.”

“It gives us a lot of flexibility and ease of management. We have all the tools in one place. We pretty much do all our storage using the Spectrum Virtualize. It makes it really easy for us to manage all our storage. It gives us the flexibility to move things in between these. I think a lot of the benefit is just the ease of use of the tool itself.”

 “There is third site replication. Right now, we're limited in our ability to migrate data between clusters. Like I said, we had to scale wide rather than tall and continue to protect our data while we migrate. Additionally, if we wanted to set up a third site for additional DR, we don't really have a good option for that.”

“The feature that's kind of missing is getting us up to the point where we can help the application owners see where their data is at, understand it, and potentially help us breakout. We've used easy tiered functions in the pools, so we're trying to help step that storage down. If they can get visibility somehow into that data, help us further break that down, or better tier and separate out their data, that would be helpful.”

“By being open source, Ceph is not tied to the whim or fortunes of any one vendor. The community of Ceph code contributors and admins is large and active. Ceph’s ability to adapt to varying types of commodity hardware affords us substantial flexibility and future-proofing.”

“In the deployment step, we need to create some config files to add Ceph functions in OpenStack modules (Nova, Cinder, Glance). It would be useful to have a tool that validates the format of the data in those files, before generating a deploy with failures.”

“Ceph lacks a little bit only in performance. It needs to scale a lot and needs very fast and well-orchestrated/configured hardware for best performance. This not a downside though, it is a challenge. Ceph only improves the given hardware.”

“StorPool is a software-only solution with practically unlimited expansion capabilities. Its performance is very high. We were able to replace our SSD-only local storage systems without any performance penalty. Its price/performance is very high!”