Posts Tagged ‘VMware’

Xsigo Data Center Fabric on VMware Solutions Exchange

Tuesday, March 20th, 2012

Visit Xsigo on the VMware Solutions Exchange and learn more about how Xsigo data center fabric complements VMware products.

vmware data center fabricThe VMware Solutions Exchange, or VSX, has quickly become a dynamic destination where you can find information about all of the VMware solutions the technology ecosystem offers.

Xsigo has partnered with VMware since 2007, and has supplied data center fabric equipment and technical assistance at numerous VMware venues, including:

  • VMworld 2009 (see the video here)
  • VMworld 2010
  • VMware Express 2010
  • VMware Executive Briefing Center Palo Alto 2012
  • VMware Executive Briefing Center Tokyo 2012

Xsigo helps you get the most from VMware with data center fabric solutions that let you dynamically connect servers to networks and storage. With a single server cable, you can deliver the bandwidth and isolated connections you need to run more VMs per server, run more application types, and deliver predictable performance.

VMware customers using Xsigo report:

  • 6X faster vMotion
  • 8X faster backup
  • Isolated connections without reliance on VLANs

Check out the VMware Solutions Exchange to see a whole range of complementary products, and learn more about Xsigo data center fabrics while you’re there.

 

 

How to Run VMware When Your Life Depends on It

Wednesday, April 20th, 2011

Bluelock offers VMware vCloud data center services. Their customers build virtual data centers at Bluelock and then use that environment as an extension of their own VMware deployment… a hybrid cloud model.

Bluelock is very good at this. They are one of only  three global VMware vCloud Datacenter service providers running certified services today. It’s safe to say they’ve thought hard about how best to run VMware. After all, it’s their life.

So we were honored to announce that Bluelock uses Xsigo virtual I/O in their VMware environment. (Read the case study here.)

Aaron Branham, director of information technology at Bluelock, cites numerous reasons why Xsigo helps their VMware environment do more. With Xsigo he achieved (compared with traditional I/O):

  • 40 times more usable bandwidth per server
  • 97.5 percent reduction in network devices
  • Eight times greater virtual machine capacity
  • 17 times more VMs per hardware device
  • 66 percent less cabling

The result of which has been very good for their business. Aaron comments, “With Xsigo virtual I/O now in place, the results speak for themselves: our clients been very impressed with the performance of their Virtual Datacenters and we’ve been able to further build our business by quickly accommodating even the most demanding workloads.”

If you are looking for a cloud provider who can dovetail seamlessly with your VMware deployment, give Bluelock a call.

If you are looking for ways to optimize your own VMware deployment, follow Bluelock’s lead and learn more about Xsigo. Here are a few resources to help:

Fuji TV Virtualizes with Xsigo

Monday, February 28th, 2011

Fuji TV, one of Japan’s largest broadcasters, has deployed Xsigo in support of their virtualized servers. Users don’t often share their architecture diagrams, but in this case they did, giving you a look at exactly what they configured (you can click on the image to expand it).

Deployed by Hitachi, this configuration  supports a range of back office applications. For details, the press release is here.

Click on the image to expand

New England Biolabs Deploys Xsigo

Tuesday, November 23rd, 2010

Biotech firm deploys Xsigo in support of their back office applications.

Highlights of the deployment include:

  • $30,000 capex savings per IBM BladeCenter deployed
  • Server failure recovery in minutes, not hours
  • Improved performance on Vmotion and backup

Taking the Worry out of Virtualization

Friday, November 12th, 2010

“The Scary Side of Virtualization.” That’s the title of Computerworld’s November 8 cover story that exposed a few “gotchas” to be aware of when virtualizing servers.

Is virtualization scary? Well, the article discusses several concerns, one of which is the co-mingling of data that should otherwise reside on physically separate networks.

That data must be protected. VLANs provide one way to maintain network separation, but Gartner analyst Neil MacDonald states in the article that “VLANs and router-based access controls alone are not sufficient for security separation.”

Yikes! No wonder one user, Jai Chanini of Rent-a-Center, comments that “for security reasons” his shop does not virtualize apps such as ERP, DB, or email.

Virtual I/O Eliminates the Fear Factor

Fortunately, virtual I/O provides a solution that retains the network isolation of a physical connection, without sacrificing real-time management flexibility. It lets you create isolated “virtual” connections that function exactly like a cable from the server to a specific network.

Here’s how it works.

  1. Create a vNIC in a server: The hypervisor views that vNIC as a standalone entity, just like a physical card.
  2. Create a vitual switch that is connected to that vNIC: In this example, virtual machine “A” sees only data from vNIC “A”.
  3. Associate that vNIC with a port on the I/O Director: Data from that vNIC then goes to that port and that port only.
  4. Connect a network to that port: In this example, data from Network “A” goes only to VM “A”. It has no path to VM “B”.

Even though VM A and VM B share a physical path from the server to the I/O Director, data from those VMs is available only at the assigned ports. It cannot be snooped on other ports, and there is no reliance on VLANs.

Isolation is the same as a physical link, but a lot more flexible. Should a server goes down, or if you want to move an application from one server to another, the virtual NICs and HBAs can be moved from that server to another device… and you still retain the original isolation.

If this sounds like something that could help you sleep at night, drop me a note and I’ll send you a white paper about exactly how this network isolation works. We have numerous customers who have virtualized their most mission-critical apps. And some of those guys are among the most security-conscious organizations in the world.  If they’re not scared by this, you shouldn’t be either.

The End of the InfiniBand Debate, or Just the Beginning?

Thursday, October 28th, 2010

Oracle yesterday bought a 10.2% stake in InfiniBand (IB) supplier Mellanox, an investment that publicly affirms the vitality of InfiniBand. IB’s success has not been overlooked by the stock market — Mellanox stock is up 200% over the past two years (vs. 50% for Cisco, for example). But strangely, IB’s growth has been overlooked by technologists, some of whom have predicted its demise. The debate over IB has been long, and with Oracle in the game the debate may now grow louder.

So why would Oracle make this investment?

They would clearly do this only if IB was a great fit for their vision for the future of computing.

The Oracle vision is naturally built around Oracle: one where all applications run fluidly across homogeneous groups of Sun servers. But aside from being closed and single-vendor centric, this vision isn’t all that different from where everyone else is going. A private cloud, after all, consists of apps running fluidly above a hypervisor on similarly-built x86 servers.

So what is it about IB that makes it a good fit in this world? Here are four reasons:

1) IB delivers substantially more throughput

40Gb solutions have been shipping for almost two years, meaning 4X the performance of 10G is available, mature and cost-effective. A single Nehalem-based server can push over 20G of I/O — Xsigo demonstrated at VMworld 2009 that — so the need is real.  Furthermore, the IB roadmap extends beyond 300G.

2) IB is ideal for high-transaction environments

IB latency is measured in nanoseconds, not microseconds as with Ethernet.

3) IB complements virtualization

When you run 20, 30, or 50 applications on a server, it is bound to drive bandwidth demand, especially when those apps might include Oracle, SQL, and Exchange. VMware identifies bandwidth capacity as a key element of  successful virtualization projects (view video here) and points out how IB capacity provides advantages over 10G Ethernet.

4) IB enables the cloud

A cloud infrastructure requires that large clusters share access to common network and storage resources, which is exactly what IB was designed for. It is scalable to thousands of nodes.

Larry Ellison summed it up in Oracle’s announcement of the deal: “InfiniBand is by far the fastest and most efficient switch fabric for running enterprise data centers.”

IB is offered by virtually every server maker, and has been for some time. And now Oracle owns a piece of that business. The often contentious nature of  any discussions involving Oracle suggests that the IB debate just got some new energy. But this investment makes a strong case for technology superiority. And at the very least should finally settle the arguments about market viability.

Game Developer Digital Chocolate Enjoys Virtual I/O

Monday, October 25th, 2010

With over 15 million active users on Facebook alone, Digital Chocolate’s games keep their infrastructure very busy. Which is why Digital Chocolate has deployed Xsigo to support at portion of that traffic.

Founded by video game pioneer Trip Hawkins (he also founded EA), Digital Chocolate boasts some hot titles including Millionaire City, MMA Pro Fighter, Rollercoaster Rush, and Tower Bloxx.

Their data center architecture is built in a cloud architecture to support the development and production of these highly popular games.

In their private cloud, they now use Xsigo to help accelerate processes and get more from their servers.

According to their team, they gained three important benefits with Xsigo:

  • Performance: MySQL servers can now support 35% more requests per second than before
  • Fast management: Virtual machine migrations now take from two to ten seconds, verses as much as 30 minutes with 1G Ethernet
  • Agility: I/O configuration changes can be completed in seconds without having to touch the servers

The latter point is significant for them.  Development and production networks are kept physically isolated in their environment (which also includes VMware ESX 4.0, Dell servers, and Compellent storage). With virtual I/O, they can move a server from one to the other in seconds, entirely in software, without sacrificing any of the isolation that separate networks provide. That can be a real time saver when putting new code into production.

The team at Digital Chocolate was initially interested in virtual I/O for the cost savings: fewer cards, cables and switch ports. What they quickly found, though, is that the operational benefits can be a lot of fun, too!

Why VMware Uses Xsigo Virtual I/O

Friday, October 22nd, 2010

Tim Myers, Sr. Architect at VMware, sat down to discuss why VMware uses Xsigo virtual I/O. Topics covered include:

  • How and why VMware uses Xsigo virtual I/O
  • Why virtual I/O is needed for the cloud
  • How virtual I/O supports VDI
  • Why VMware trusts Xsigo virtual I/O

For more on virtual I/O and the cloud, download the Yankee Group white paper, How Ethernet-Based Virtual I/O Paves the Way for Cloud Computing, here.

Newedge Group Deploys Xsigo

Wednesday, October 20th, 2010

Newedge Group, a global financial services firm, builds a cloud with Xsigo in their 2000 server production environment.

A View to the Future

Friday, February 19th, 2010

VMware last week unveiled the VMware Express, a rolling demo platform largely dedicated to VMware View, their virtual desktop environment. Essentially a powerful data center-on-wheels, the truck will provide a demonstration environment at events across North America during its year-long tour.

busXsigo was selected to provide virtual I/O on the VMware Express for the obvious reason: virtual I/O significantly increases compute density. When space and power are at a premium, you need virtual I/O to make the most of them.

Jon Owings of Veristor captured in a recent blog post several points about why Xsigo virtual I/O and View are so complementary.

Starting with a VMware View reference architecture (shown below), he figured out how it would be implemented with traditional I/O. Jon comments, “It really struck me how many I/O connections (Network or Storage) it would take to run this POD. Minimum (in my opinion) would be 6 cables per host with ten 8 host clusters that is 480 cables!”

Next, he did the same math with Xsigo and arrived at 160 cables, a 66% reduction.

ViewPOD5k

To further his point, he calculated the total server I/O bandwidth available with traditional I/O: 960Gbps was the total.

For Xsigo the total bandwidth came to 3,200 Gbps, over 3X more!

Jon surmises that a cost reduction would result from that reduced complexity (he’s right), and concludes, “Being in the data center I like any solution that makes provisioning servers easier, takes less cabling, and gives me unbelievable bandwidth. So just in the way VMware changed the way we think about the data center, virtual IO will once again change how we deal with our deployments.”

If you have chance to see the VMware Express as it tours the country, check it out. You’ll see not only the future of desktops, but of enterprise data centers as well.