Not a big surprise, but VMworld was overwhelmingly focused on software. I was concerned that great hardware might get lost in the software noise, so I looked for hardware standouts using two criteria:
- They had to have some relevance to virtualization. Most offerings could check this box – we were at WMworld, after all.
- They had to have some potential impact on Infrastructure & Operations (I&O) for a wide range of companies.
So here are my two winners that I’d advise my fellow I&O hardware aficionados to watch:
NEC FT x86 servers: FT VMs the Easy Way
VMware has a Fault Tolerant (FT) VM product, and it works. But with major compromises, among them the need for a dedicated 10G network link to keep the FT pairs in sync, a requirement for a management system, additional license costs, and a limitation to a single-core VM. If you require FT VMs and are willing to place your bet on a monolithic FT server in lieu of distributed software FT, think about running your standard VMs under your standard VMware license on a fault-tolerant server from NECor Stratus (identical hardware, different channel and support models). These servers look to the VMware software and management environment exactly like a standard Xeon-based dual-socket server with up to 12 cores total. However, under the covers they are a lockstep hardware FT system, the latest iteration of the hardware that NEC and Stratus have been selling for years. What is different is that they are fully certified and supported running VMware.
The overall reliability of these systems, as reported publicly on Stratus’s website, is well in excess of the “5 nines” availability bar set for many FT environments, and does not require any additional software or application modification. The comparison of all the financial and operational elements will be time consuming, but at first glance, factoring in license costs, additional hardware costs and performance along with the potential for more than single-core VMs, an FT-based hardware solution may make sense if you do not absolutely need the ability to continue operations on a physically remote system. For me, the simplicity was compelling – take the VMs you already have, run them on the FT hardware, and then ignore them. VMware will undoubtedly improve their FT offering, but for now an FT platform running standard VMs may be an alternative worth considering. Consider either NEC or Stratus, depending on your vendor relations and desired style of support.
Xsigo I/O virtualization: Converged Infrastructure for the Masses
Virtualized I/O has begun to emerge as an element of advanced integrated infrastructure solutions from Cisco, Dell, HP and IBM, but has not yet emerged as an offering for mainstream rack servers or as an add-on technology for existing servers. XsigoSystems, a small privately held company, was at VMworld showing their I/O director technology. Xsigo’s I/O Director at first glance looks like it delivers a subset of HP Virtual Connect or Cisco UCS I/O virtualization capability in a fashion that can be consumed by legacy rack-mount servers from any vendor. I/O Director connects to the server with one or more 10 G Ethernet links, and then splits traffic out into enterprise Ethernet and FC networks. On the server side, the applications, including VMware, see multiple virtual NICs and HBAs courtesy of Xsigo’s proprietary virtual NIC driver.
Controlled via Xsigo’s management console, the server MAC and WWNs can be programmed, and the servers can now connect to multiple external networks with fewer cables and substantially lower costs for NIC and HBA hardware. Virtualized I/O is one of the major transformative developments in emerging server technology, and I will be examining it in more detail in future research.