The Hidden Gem That Is StorPool

Chris Evans knows his storage. After seeing StorPool present at Storage Field Day, he considers them a hidden gem. They offer a high-performance scale-out block storage solution running on commodity hardware and Linux. Using their self-developed on disk-formats, inter-node protocols and cluster management, StorPool was able to offer the same performance as Microsoft’s HCI specification, without the need for a costly NVDIMM cache or RDMA. Chris thinks they have a very interesting solution for the right customer.


Rubrik, Cohesity and the Battle for NoSQL Backup

As always, Chris Evans has a keen insight into the backup market. He recently updated a post about Rubrik’s acquisition of Datos IO to add in some additional perspective on the NoSQL backup market. With Cohesity acquiring Imanis Data, Chris sees this as both companies trying to bridge the gap between traditional backup and data management, something always easier said than done.


You’re Wrong About Data Protection Policy

This episode of the On-Premise IT Roundtable was recorded at Tech Field Day Extra at Dell Technologies and looks at data protection policies. The group discusses why you’re probably doing it wrong. Essentially business and compliance needs are not top of mind when these policies are established. They further discuss how vendor defaults and application owners add to the confusion.


Can the WekaIO Matrix File System Be Faster Than DAS?

Direct Attached Storage (DAS), offers a lot of advantages in terms of speed and latency, but comes with a suite of problems. If nothing else, it forces applications to deal with data integrity and resiliency issues instead of having it as part of SAN. Chris Evans looks at how WekaIO addresses the shortcomings of DAS while keeping the high-performance and low latency. At Storage Field Day last month, they showed how their scale-out parallel file system can provide exceptional performance that’s only limited by the bandwidth of the network card. By rewritting at the file system layer and lower, WekaIO is able to offer distributed storage built to take full advantage of modern storage media.


#91 – Storage Field Day 18 in Review

In this episode of the Storage Unpacked podcast, Chris Evans and Martin Glassborow discuss what happened at Storage Field Day last month. The companies pretty cleanly divided between scale-out primary storage and data protection solutions. They touch on all the presenters, and where listeners can learn more about the event. Be sure to give it a listen as a preview before watching all of the event presentation video on our site.


VAST Data Launches With New Scale-Out Storage Platform

As one of the delegates at Storage Field Day, Chris Evans got to see VAST Data come out of stealth at the event. The company offers a new storage platform built on a disaggregated shared-everything architecture. Using a combination of QLC NAND flash and storage-class memory in an enclosure, linked across an NVMe fabric, any storage controller can talk to any NVMe device on the fabric. The result is a scalable architecture with no inherent pinch points, that could scale to thousands of storage nodes and tens of thousands of controllers.


Exploiting Secondary Data With NDAS From NetApp

Chris Evans got to hear about NetApp Data Availability Services at Storage Field Day last month. While the service is being pitched as a tool of IT generalists, Chris sees longer term value for it as a secondary option for data search and analytics. If this is a broader data management play, the company will need to expand beyond structured data, but Chris was impressed by what he saw.


Podcast #3 – Chris & Matt Review the SFD18 Presenters

In this podcast episode, Chris Evan and Matt Leib did an off the cuff rundown of the Storage Field Day presenters. The event was jam packed with interesting IT companies, and the two do a great job of running through what stood out from Cohesity, Datera, IBM, NetApp, StorPool, VAST Data WekaIO and Western Digital.


MAX Data: Optimising Application Performance With SCM

Chris Evans digs into NetApp’s MAX Data, a tiered storage solution that comes from the company’s acquisition of Plexistor back in 2017. Chris sees this as a must watch technology for 2019, even if he still needs to see some solid performance data before making a final judgement. To learn more about how MAX Data can “really exploit the potential of SCM and technologies like Intel Optane”, be sure to watch their presentation from Tech Field Day at NetApp Insight.


NetApp buys into Kubernetes Orchestration with StackPointCloud

Chris Evans breaks down the reasons behind NetApp’s recent acquisition of StackPointCloud. The company gained a management Kubernetes provider that works across clouds. Now they are using it to run their own NetApp Kubernetes service. This is another vital step for the company from transforming from selling storage hardware and software into a storage service provider.


Elastifile Cloud File Service Now Available on Google Cloud

Chris Evans looks at the significance of Elastifile now offering a managed scale-out file service for Google Cloud, calling it nothing less than a coup for the company. He goes into the background on Elastifile’s Cloud File Service, which he first saw at Storage Field Day in 2017. By being a native first-party service with GCP, Elastifile gets the benefits of using native APIs, have billing done through Google, and have Google manage their scaling. Having native integration on GCP makes Elastifile “a winner” for Chris.


Western Digital Redefines DRAM Caching

Chris Evans digis into Western Digital’s Ultrastar DC ME200. This is a solution to extend server DRAM using caching into byte-addressable NAND. Chris does a great job of giving this a technical overview, and considering if this will find a welcoming market, or simply cover up Western Digital’s lack of a true storage-class memory solution. He also highlights similar efforts to use NAND as a memory cache, specifically citing Diablo Technologies’ Memory1 solution, which he saw back in 2016 at a Tech Field Day event.


HDD Capacity Threshold Reaches 15TB

Chris Evans highlights the new king of HDD capacity, the 15TB Ultrastar DC HC620 from Western Digital. While only representing a 7% increase in capacity from 14TB drives, the real innovation here is these drives use host-managed Shingled Magnetic Recording, as opposed to being managed directly on the drives. This has big implications for hyperscalers, who can see greatly improved performance. Dropbox is one of the initial customers to use the drives, and Chris got a deep dive into their “Magic Pocket” architecture at Storage Field Day earlier this year.


Scality Zenko Product Brief

In this paper, Chris Evans outlines Scality’s Zenko platform. This acts as a storage controller across object stores, whether on-prem or in the cloud. This draws on much of what Chris saw from Scality at Storage Field Day last year.


Enterprise-Class Public Cloud

Chris Evans takes a look at what “enterprise-class” cloud computing would look like, which he heard about extensively from Oracle at Cloud Field Day earlier this year. Oracle’s Cloud Infrastructure provides a more managed environment than other public clouds, making it well suited for enterprises that can’t completely rewrite applications. For Chris, there is definitely a market for a cloud provider that doesn’t strictly take a hyperscaler approach.


What Next for Violin Systems?

Chris Evans takes a look at the current state of Violin Systems, a company he got to see at Storage Field Day back in 2016 when it was known as Violin Memory. The company has gone through more than just a name change since that time, having significant financial and management shakeups. Chris overviews what made them an early all-flash storage leader, and where he sees their technology taking the company going forward.


Pure Storage AIRI – CI for AI

Pure Storage’s AIRI platform provides a converged infrastructure approach for ML and AI workloads, combining FlashBlade and NVIDIA DGX-1 servers. After looking at the Storage Field Day Exclusive videos from Pure Accelerate in 2017, Chris Evans sees this as an important step for FlashBlade, what he hopes is a first in a series of reference architecture solutions for specific applications.


Why Deterministic Storage Performance is Important

Chris Evans looks at why deterministic performance is important in storage. This effects everyone from hyperscalers, to smaller enterprises using HCI. Chris recalls SNIA’s Storage Field Day presentation from last year discuss how the organization is addressing tail latencies as part of this issue. Regardless of the size of an organization, consistent average performance is good, but significant poor performing outliers can be a significant challenge.


The Britpop Battle of Rubrik and Cohesity

The rivalries between data protection companies remind Chris Evans of the glory days of 90s Britpop. Instead of Blur and Oasis, for him the rivals of this new vanguard are Rubrik and Cohesity. He sees both as working to centralize data protection to offer it as a service, with the ultimate goal of fending off cloud-based solutions.


PowerMax, VMAX, XtremIO – the Agony of Choice

Chris Evans looks at Dell EMC’s newly announced PowerMax, an all-NVMe storage array. When fully configured, PowerMax can hit up to 10 million IOPS at under 300 microseconds of latency, while still being able to access the mature data service of their VMAX line. Chris raises a similar to point to Dell EMC’s XtremeIO presentation at Storage Field Day last year, that with many families of storage arrays, Dell EMC needs to be very clear which line is best for prospective customers.