VM-Net

On March 29th HP announced a number of new and enhanced products in its G6 ProLiant server and blade range. The new G6 servers and blades announced utilise Intels new Xeon 5600 Series of processors code named Westmere. Whilst the Westmere CPU’s are available with 4 cores its the  six-core (12x hyperthread) Westmere’s that will provide a significant gain over 4-core Gainestwown 55xx in virtualization environment. The new G7 servers announced  are alledged to offer upto 97% reduction in power and cooling costs, 27 time improvment in perfomace per watt whilst supporting higher consolidation ratios.

The new  G6 servers include:

• 4 x New ProLiant  blades
• 1 xNew workstation blade
• 6x  New DL servers
• 2 x New Ml server
• 3 x Scale out server

The new G7 servers include:

• HP Proliant DL165 & SL 165z Server
• HP ProLiant DL385

VMotion is one of those features of  VSphere, and of VMware infrastructure 3, that is quite simply amazing the first time you watch it move a live vm from one ESX server to another with any disruption. However to achieve this magical feat VMotion requires a network link of at least LAN speed to complete the task.

As cool as this is, one question often asked is, “How do we take that one step further, and perform VMotion between datacentres?” This, of course, is a non-trivial thing to do.  There is the challenge of moving a VM over distance (which involves some degree of additional latency) without dropping sessions. To maintain sessions with existing technologies means stretching the L2 domain between the sites, not pretty from a network architecture standpoint. And then there is the storage piece. If you move the VM, it has to remotely access its disk in the other site until a Storage VMotion occurs.  

Last year, Cisco and VMware began the task of trying to solve these long distance VMotion issues with the target of seamlessly migrating a VM between two  separated by a reasonable distance. The joint Cisco/VMware lab in San Jose has run number of tests over varying distances (simulated with reels of optic fiber) as a proof of concept.

This was demonstrated at Cisco Live in San Francisco. The demo as it stood incorporates a distance of 80km (50 miles). See above the above diagram.

This proof of concept is aimed at the following requirements:

1. Load balance compute power over multiple sites: Migrate VMs between datacentres to “follow the sun”  or to simply load balance over multiple sites. Enterprises with multiple sites can also conserve power and cooling by dynamically consolidating VMs to fewer datacenters (automated by VMware Dynamic Power Management (DPM))—another enabler for the Green datacenter of the future.
2. Avoid downtime during DC maintenance:applications on a server or datacenter infrastructure requiring maintenance can be migrated offsite without downtime.
3. Disaster Avoidance: Data centers in the path of natural calamities (e.g. hurricanes) can proactively migrate the mission critical application environment over to another data center.

Use cases #2 and #3 above also require a Storage VMotion to move the disk image to the alternative datacentre.

HP yesterday announced 11 new additions to their ProLiant server platform including new rack, tower and blade models. Timed to coincide with Intels “Nehalem EP” Xeon 5500 processor launch these 11 new servers create 1,000 new server SKU’s. 

The new G6 blade servers that suppport the new Nehalem chips include the ProLiant BL280c, the BL460c, and the BL490c. With Intels “Nehalem EX” octocore Xeons coming later this year for four-socket and larger servers as well as the six-core “Istanbul” Opteron chips from AMD it looks like even more ProLiant server SKU’s are on the cards.  As the names suggest, they are all compatible with the SMB-style “Shorty” c3000 chassis or the larger c7000 chassis. 

HP is also integrating its ThermalLogic technologies, created for its BladeSystem blade servers, into its rack-based ProLiants. This technology allows different parts of a system to be cooled independently and monitored via an array of 32 different smart sensors allowing the server to adjust fan speeds and memory and I/O capacity as workloads change.

The new G6’s  also include dynamic power capping, which will allow administrators to set a power threshold for a server and  allow the server itself to quiesce various components to ensure that it stays under the level set. The technology will also allow one or more machines in a rack to run hot if their workloads require. Another green change within the ProLiant G6 family is the move to to a standard set of power supplies. HP is offering a 460-watt supply with a 92 per cent efficiency, a 750-watt supply with a 92 per cent efficiency, and a 1,200 watt supply with a 90 per cent efficiency allowing administrators to  further reduce their power consumption. 

Along with the G6 launch HP have also released a new set of SmartArray disk controllers that offer up to a 200 per cent improvement in I/O operations per second, and thanks to the Nehalem processors G6 servers will offer at least twice the memory capacity than their predecessors. In many cases, the G6 servers will have 18 DDR3 DIMM slots on the larger servers, on the smaller machines they will have 12 DIMMs.

According to HP they have packaged the industry’s best ideas, innovations and components and put them through 11,000+ quality and design tests and built the new G6’s on the experience of more than 17 million ProLiant servers and 20 years of industry leadership. You can register for HP’s Virtual Web Jam Event  on April 7th to hear more about their latest G6 products: www.hp.com/go/proliant-jam.