Hitachi Pulls Itself Together In The Datacenter
May 10, 2018 Jeffrey Burt
Hitachi is a massive multi-national conglomerate that has more than 300,000 employees and 950 subsidiaries and a reach that extends into a wide array of industries, from aircraft and automotive systems to telecommunications, construction, defense and financial services. It also is among the world’s largest IT companies, nestled in there among the likes of Apple, Amazon, Microsoft, Google, Samsung. Hitachi’s sprawling technology capabilities ranges from compute and storage appliances in its well-known Hitachi Data Systems (HDS) unit to datacenter management software, data management and business intelligence, and the Internet of Things.
For the past several years, the company has worked to determine how best to take these siloed capabilities and integrate them to take advantage of the rapid changes in the industry around data, distributed computing, the cloud and, more recently, emerging technologies like artificial intelligence, machine learning, and analytics.
The company has not only developed systems and devices for all these industries, but also technologies that have allowed them to communicate with each other and other systems and to collect and storage massive amounts of data, giving them levels of intelligence and connectivity before the IoT was a thing itself. In 2016, Hitachi looked to take advantage of the dozens of industry-specific IoT solutions and more than 16,000 employees whose work was IoT-related and pulled them all into an umbrella organization called the Hitachi Insight Group. The move included launching Lumada, an adaptable platform that not only integrated the commercial technologies from across the various product portfolios, but also included data analytics and intelligence capabilities. That came just months after HDS and Pentaho, Hitachi Group’s data integration and analytics business, partnered to develop the company’s Scale-Out Platform 400 hyperconverged infrastructure solution.
In September last year, Hitachi brought together HDS, Pentaho, and Hitachi Insight Group to create Vantara, a single business unit with some 7,000 employees aimed at addressing both the operational technology and IT needs of its lineup of commercial and industrial customers, all of whom are moving into the cloud and the IoT, reworking their infrastructures to adapt to an increasingly digital era and need capabilities like analytics and data management. The move with Vantara puts Hitachi into closer competition with the likes of Dell EMC, Hewlett Packard Enterprise, Cisco Systems, IBM, and the growing numbers of other top-tier vendors who are looking to help enterprises move into a more cloud-based, software-defined environment.
Vantara will be able to differentiate from many tier-two competitors through its broad range of offerings, Nathan Moffitt, senior director of infrastructure solutions marketing for Hitachi Vantara, tells The Next Platform.
“When it comes to the tier one players, our vision of autonomous operations is someplace that other people are going, but competitively, we’re ahead of the curve, especially in our willingness to take a more open approach to things, not just in terms of bracing other folks, but in terms of having the API-driven infrastructure in place to allow us to facilitate and accelerate that integration between different companies,” Moffitt says. “The strategic direction we want to take here is, how do we help customers modernize their overall datacenter operations so they accelerate and create opportunities for new innovation to drive the business forward? Historically, if you looked at Hitachi, we would have started with [agile data infrastructure] and maybe gone into [modern data protection], but that would be about it. But as we have matured as a company and brought these three companies together, our perspective on how we help customers modernize their datacenters has evolved as well.”
The plan is to add capabilities to what Vantara can offer throughout the year, starting this week with new all-flash and hybrid storage systems and integrated analytics and automation software that leverages AI techniques. New to the Virtual Storage Platform (VSP) series are the all-flash enterprise-level F700 and F900 (pictured below) and midrange F350 and F370. In the hybrid G series, Vantara is adding G700 and G900 at the high end and G350 and G370 midrange systems. The new arrays can scale up to more than 2.4 million IOPS and 41 GB/sec of bandwidth, more than the 1.4 million IOPS and 24 GB/sec in the high-end predecessors. Deduplication is up to 3.4 times faster and fives times the compression speed.
“With all-flash, our initially strategy was, let’s just hit the high end as hard as possible and get a good footprint in terms of flash,” explains Moffitt. “What we’re seeing from our customers and partners is they want us to bring some of that downstream. We’re going to expand that all-flash into the midrange, which allows us to go to another customer base and get a lot more aggressive against some of our competitors who focus on the midrange versus the enterprise and high-end enterprise. Our history was definitely in the high end. If you look at HDS in the past, we had a lot of success there. But it wasn’t in the midrange. We recognize there is an opportunity for us so we’re bringing these systems downstream so we can address more of that midrange space and help go after growing customers that may eventually become the next genesis of our enterprise and high-end enterprise customers.”
We don’t follow the midrange much here at The Next Platform, but clearly the vendors that do serve the midrange often then have the margins to better support the high end, and those companies that only have a high end business – think about HPC centers and hyperscalers – tend to have very cyclical and not particularly profitable businesses. Those vendors who lose this base, which is often a feeder to create larger enterprises, do so at their own risk.
At any rate, Vantara also is taking a three-step approach to NVM-Express. As we’ve noted before, the high-profile protocol designed to increase throughput and lower latency for flash and other non-volatile memory is beginning to make the move from servers to storage systems, and Hitachi is following a similar path. The company already has NVM-Express in its hyperconverged systems, with the next step address rack-scale flash.
“This is an area where we’re seeing a lot of interest because it aligns with a very specific set of use cases that can really benefit from that NVM-Express performance,” Moffitt says, adding that the company wants to address areas where it can help with workloads at a price point that organizations can bear. “After that, we’ll see NVM-Express will then move into our enterprise arrays like the VSP line. The reason we’re doing that at the third stage is, one, it’s about making sure the technology, especially on the back end, has all the resiliency capabilities in it, like dual-port drives, making sure the standard is mature, as well as making sure the price point is right. Storage used to be a huge premium offering. You’re recognizing now that you’ve got to have that priced dialed in correctly.”
Vantara also is rolling out integrated operations software that includes analytics, resource monitoring and delivery, and automation driven by AI technology. Hitachi Infrastructure Analytics Advisor (HIAA) provides predictive analytics by mining telemetry data from servers, storage appliances, networking systems and virtual machines to optimize performance, troubleshoot issues and forecast when a business may need to buy new storage systems. Hitachi Automation Director (HAD) orchestrates the delivery and management of IT resources. Enhancements to the software include integration with IT service management (ITSM) tools, including the ServiceNow platform, for better resource tracking, improved REST API integration for working with third-party resources, integration with HIAA to fix issues and provide storage configuration data, and automatically setting up the Hitachi Data Instance Director (HDID) software.
The decision to work more broadly with third-party vendors is new for Hitachi. HIAA works with a variety of other parties, including Microsoft, VMware, and OpenStack/KVM in the hypervisor space, Cisco and Brocade in networking, and HPE, IBM, Dell EMC, NetApp and Pure Storage in storage. For HAD, partners including Microsoft, VMware, Red Hat, Oracle, OpenStack, Brocade and Cisco. The SVOS now supports container platforms based on Kubernetes, Docker, or OpenStack.
“Everything we have is API-based so that we can draw information in from other sources to create a more intelligent solution, but we can also pass information out if we’re not the master in the environment so we can make other things smarter,” Moffitt says. “We are taking a very vendor-neutral approach here and that is something that is very new to us. The integration of HIAA, the analytics piece, and the HAD, the automation piece, plus the integration of pulling in more vendors externally that we can integrate with to build up the solution is new. The products grew out of being much more storage-focused, so we recognized that there is a much broader opportunity here and we have to be willing to integrate with partners. It’s a huge thing.”