Cisco and Microsoft this week announced joint reference designs for data warehouses and virtual servers, officially combining the Cisco Unified Computing System (UCS) hardware with Microsoft's Hyper-V hypervisor and SQL Server for the first time.
Aimed at medium to large enterprises, the SQL Server and Hyper-V designs are separate, with SQL running on bare metal and Hyper-V able to handle a variety of applications. The reference designs also involve storage partners, with NetApp in both and EMC in the new SQL Server offering.
"They key word here is choice," said Cisco's data center marketing manager Rex Backman in an interview. The Hyper-V and NetApp announcement marks a new direction for Cisco's server virtualization partnerships, which had previously been focused on vBlocks--bundles of a Cisco server with EMC storage and a VMware hypervisor.
EMC's involvement in the SQL Server announcement shows that its partnership with Cisco is still going strong, but the addition of a Microsoft and NetApp virtualization option will help Cisco counter criticism that its offerings are too proprietary. Almost every other switch vendor's pitch involves openness, with Brocade two weeks ago announcing plans for what it calls Virtual Compute Blocks--bundles of sever, networking, and storage that compete directly with vBlocks and with the new reference designs.
Cisco's emphasis on choice even extends into switching: The reference design for data warehousing mandates a Cisco UCS server, Microsoft SQL Server 2008 R2 software, and NetApp or EMC storage, but only suggests a Cisco Nexus switch to link them together. However, the configuration is optimized for the Cisco switch, with performance figures based on it. "We're going to suggest the Nexus; it's in our nature," said Backman.
According to Cisco, the company's engineers have been testing UCS with Microsoft Hyper-V since it entered the server business, but this is the first time that the two are officially validated to work together. The reference designs that combine UCS with SQL Server are part of Microsoft's SQL Server Fast Track Data Warehouse program, which also includes configurations running on hardware from HP, Dell, Bull, EMC, and IBM.
So why run Microsoft on a Cisco server? "We've done a lot to develop a more efficient, more flexible, more agile server platform," said Backman. "There's our architecture around extended memory for SQL applications, as well as unique things with Microsoft around manageability." The Microsoft-specific enhancements include integration with Powershell, which lets administrators access Cisco switch and server functionality through the Windows command line shell and scripting language.
Extended Memory is a Cisco technology that uses custom application-specific integrated circuits between a server's memory sockets and processor to cram more RAM into each physical server--two to four times the amount offered by competitors, according to Cisco, currently scaling up to 1 TB. A large and low latency memory is important for both SQL Server and Hyper-V, letting the former cache large databases and the latter quickly switch between virtual machines. Though the technology is Cisco-specific, it's transparent to software and other hardware components, meaning it doesn't require any other Cisco gear. "It uses standard DIMMs so it's very cost effective," said Backman.
Competing on cost is a departure for Cisco, but, as with the emphasis on openness, it is likely a response to its critics. In particular, Cisco may be reacting to a Gartner report that said customers could save money by using another networking vendor, even when the costs of managing a multi-vendor network are taken into account. Cisco's previous response has been to emphasize that the enterprise network can be a way to add value to an organization, not just a cost center.
Servers running VMs are still just servers. Stick with security practices you know to minimize risk. Also in the new all-digital InformationWeek supplement: Trust but verify when it comes to hypervisor security. Download it now. (Free registration required.)
Subsea cable alternatives can provide a level of comfort for those concerned about disruptions. Realistically, they will continue to play a small, but critical role in the short term.
Maintaining existing network infrastructures often incurs huge technical debt before newer technologies are adopted over time.
Amid ongoing regional uncertainties, telecom infrastructure in the Middle East has expand to include terrestrial low-latency network infrastructure, which offers resilience and agility in navigating political complexities without relying solely on undersea networks.
