Enterprise Storage Reviews
Enterprise disk, hybrid and all flash array (AFA) reviews that cater to midmarket enterprises and up.
by StorageReview Enterprise Lab

Violin Systems FSP 7650 Review

Violin Systems isn't exactly a new company; we have been covering them for six years now. The company began as a pioneer in the all-flash arena and ran into a few woes after going public. However, the company has risen from the ashes with new investors backing the firm and has new deployment options along with the same cadre of high performing gear. We've been working with Violin in the lab for some time now; today we are looking at the 7650 array, which is the “extreme performance” all-flash model in the Violin Flash Storage Platform (FSP). 

Read more
by StorageReview Enterprise Lab

Toshiba KumoScale NVMe-oF (Newisys NSS-1160G-2N) Review

At the Open Compute Project this year, Toshiba announced the release of its new software around NVMe over Fabrics (NVMe-oF), KumoScale. KumoScale is designed to maximize the performance benefits of direct-attached NVMe drives over the data center network through disaggregation, abstraction and management integration. The software brings the already high-performing NVMe SSDs by allowing diskless compute nodes (with only a boot drive) to access this flash storage over high-speed fabrics network connectivity. This type of connection will bring the networked storage to near-peak performance. 

Read more
by Brian Beeler

Dell EMC XtremIO Replication Review

Today Dell EMC has announced the addition of native asynchronous replication in XIOS 6.1 for XtremIO all-flash storage arrays. Until now, replication for XtremIO has been handled by software tools like Dell EMC RecoverPoint for snapshot-based replication or Site Recovery Manager for VMware shops. There're also the options of pairing XtremIO and Dell EMC VPLEX for a variety of replication use cases. While these options solve the problem for many customers, they can also introduce increased complexity and cost, depending on the hardware in the environment. As such, a native metadata-aware replication solution for XtremIO has always been on the radar and is the number one feature XtremIO customers have been asking for. 

Read more
by Adam Armstrong

Fujitsu Storage ETERNUS AF250 S2 Review

Earlier this year Fujitsu announced updates to its ETERNUS all-flash (AF) line, with the introducing of AF250 S2 And AF650 S2. Having previously looked at the ETERNUS AF250 and found it to be a strong performer (winning our Editor’s Choice award) especially compared to its pricing. Now we are taking a good look at the updated version to see what improvements have been made and how they stack up.

Read more
by StorageReview Enterprise Lab

Reduxio HX550 Review

The Reduxio HX550 is a dual controller hybrid storage array within a 2U chassis. The system frontends sixteen 7.2k 2TB hard drives with eight 800GB SSDs to create a 38.4TB raw data pool in each system. Stopping there makes the HX550 sound just about like any other midrange hybrid offering that's trying to appeal to the needs of the SMB and enterprise markets. However, Reduxio is much more, in that it offers several unique twists that are uncommon or even entirely novel in this class of storage. Primary amongst these is Reduxio's "raison d'être," which is that data should be protected at all times. Reduxio executes on this mission by offering BackDating, which offers one second RTO and RPO, effectively letting an organization roll back at second granularity at any time. Reduxio customers have found this methodology to be effective at dealing with all sorts of data loss or access issues, especially in the face of the burgeoning issue of ransomware. 

Read more
by StorageReview Enterprise Lab

Oracle Cloud Infrastructure Compute Bare Metal Instances Review

The Oracle Cloud Infrastructure includes a wide variety of service offerings including compute, storage, networking, database, and load balancing--in effect, all the infrastructure necessary to construct a cloud-based data center. In the context of this review, we're interested in the Oracle Cloud Infrastructure Compute category, with a very specific focus on their bare metal instances. Like most cloud providers, Oracle offers virtualized compute instances, but unlike most other virtual offerings, Oracle can back these with shapes that contain up to ~25TB of NVMe storage for applications that need low latency. As great as those are, Oracle has further upped the ante for cloud compute performance by offering the industry's first high-performance bare metal instances, ideal for mission-critical applications where latency is paramount. Instances come with up to 52 OCPUs, 768GB RAM, dual 25GbE NICs and up to 51TB of locally attached NVMe storage. For those that desire more, up to 512TB of network-attached NVMe block storage is available as well as GPU options. All the different Oracle compute offerings run on a highly-optimized software-defined network tuned to minimize contention and maximize performance.

Read more
by Brian Beeler

NetApp AFF A200 VMmark 3 Results Published

At the time of our initial NetApp AFF A200 review publishing, we were in the the process of finalizing our VMmark 3 test process and working through our first submission. We've since streamlined the VMmark 3 submission process and our audited submissions are approved at a regular pace. The recent approvals include the results from the NetApp AFF A200. What makes these results particularly interesting is that we have data with both data reduction services enabled and disabled. For most arrays enabling these services kneecaps performance; not so with the A200.

Read more
by StorageReview Enterprise Lab

Dell EMC Unity 450F All-Flash Storage Review

The Dell EMC Unity family of storage is designed to meet the needs of the midrange customer with a variety of options including hybrid, all-flash and VSA deployment options. Dell EMC Unity storage generally takes over for the previous EMC VNX family. To say it's an evolution of VNX, though, isn't entirely accurate; EMC completely revamped the interface and continued ease-of-use and deployment developments that largely got underway with the VNXe in 2014. The interface is totally new, and the underlying code has been updated as well. The Unity array consists of a 2U chassis with dual-active/active controllers powered by Intel Xeon processors and 25 2.5" drive bays (there is also an option for 12 3.5" bays for the hybrid version). Systems scale up for more capacity and Dell EMC leverages TLC flash in the SSDs to help drive affordability. The systems support block, file, and VMware VVols with concurrent support for native NAS, iSCSI, and Fibre Channel protocols. The all-flash systems (or hybrid with an all-flash pool) include data reduction features for further TCO gains. All Dell EMC Unity arrays can tier to the cloud easily and offer cloud-based analytics via CloudIQ with no additional investment. 

Read more
by Mark Kidd

Premio FlacheSAN1N4C-D4 Flash Array Data Transfer Node in SCinet

In November we attended the 2017 International Conference for High Performance Computing, Networking, Storage, and Analysis -- better known as the 2017 Supercomputing Conference (SC17). Each year at the Supercomputing Conference, researchers and vendors from academia and industry assemble SCinet: a supercomputer and network fabric that serves as testbed and proof-of-concept for new high performance compute, networking, and storage technologies. Premio is one of the vendors who contributed hardware and expertise to the 2017 SCinet in the form of two FlacheSAN1N4C-D4 data node enclosures, with each node benchmarking  up to 24GB/s (168Gbps) sustained read throughput and up to 3.6M IOPS. This allowed SCinet to have the capacity to transfer large datasets at a fast cache speed.  

Read more
by StorageReview Enterprise Lab

Datrium DVX with Flash End-to-End Review

Datrium delivers "open convergence," which is what they call the next generation of converged infrastructure. There may be as many riffs on convergence as there are vendors in the enterprise IT space, and for their part, Datrium isn't shy about promoting their vision of what infrastructure can look like. Datrium's view sees compute, primary storage, secondary storage and cloud coming together in a highly resilient configuration that is scalable and easy to manage without a silo for each class of storage. Furthermore, because most data calls will hit compute nodes with on-board flash cache, Datrium can deliver tremendous performance without having to go to the data nodes in nearly all cases. This translates to 200GB/s and 16GB/s in peak 32K read and write bandwidth and 18M IOPS 4K random read.

Read more