If you’ll permit the grandiosity, I believe that virtualization is the best thing since the invention of sliced bread (probably even better than that, actually). I say this because Fandotech has been in the throes of virtualizing much of itself recently and is seeing the benefits of this effort. When I say we’ve been “virtualizing,” I should probably tell you what I’m talking about and how this can benefit you. Here’s the breakdown:

The Old Way
Let’s discuss a hypothetical typical client that hosts an ecommerce website on their own servers (not shared with other clients). To do this the old way, they’d need one or more physical servers.

Typically, the website itself will be hosted on one physical server that is connected to the Internet. That server also has a “backdoor” network connection to a second server hosting the database. This second server is not connected directly to the Internet, thus offering greater security for the database. Two servers in this example so far, but there might be more for various reasons. Each server is chewing up resources, has an ugly carbon footprint, and costs the client real dollars. Each server also takes up physical space and is turned on and operating 24/7/365. And, sadly, each server is probably not engaged to its capacity, so resources are wasted as these servers idle away the days, months, and years.

The New Way
There’s now a better and far more efficient way to use resources and still securely host an ecommerce website, likely with improved performance and other advantages. Our client will need one or more virtual servers.

A virtual server is all the data from the physical server’s hard drive. For instance, it’s everything the old server was except for the hardware. In our example, the ecommerce client will need two of these virtual servers, one for the website and another for the database.

These virtual servers will be housed on a host server. But, from this point forward, the rest of the infrastructure is our responsibility. We do all of this already, so the client can just “move in.”

Host servers are highly capacious, beefy, herculean servers that can host numerous virtual servers. They have the hard drive space, memory capacity, and processing power to operate the functions of multiple virtual servers, with each virtual server being walled off from the others (each operates in a virtual shell, as if the other “servers” weren’t there).

Two or more host servers that are linked together are called a cluster. Clusters provide enhanced security, performance, and data redundancy. In a cluster, one host server can back up another, or it can provide additional hosting space. Basically, the cluster increases capacity and provides redundancy in case of hardware failure.

The Benefits
At the time of this writing, we’ve migrated 42 of our servers onto a cluster of 5 virtual servers. Those old servers? Unplugged and recycled (or repurposed for worthy causes). The resources needed to run 5 host servers are far less than those needed to run 42 typical physical servers. The carbon footprint, impressively less. Those five powerful host servers are able to deliver improved performance over the 42 old servers of various makes, models, and horsepower. And if performance begins to lag, another host server can be added to the cluster to expand its overall capacity and relieve any bottlenecks.

What about our hypothetical ecommerce client? They benefit from a smaller monthly hosting fee for two virtual servers versus two physical servers. They sleep better at night (we imagine), knowing they’re helping to save the environment. They can take advantage of swift, easy, and relatively cheap expansion possibilities. And they likely enjoy faster performance and no hardware maintenance woes.

Does all that sound good? If so, contact your account representative to discuss how virtualization can benefit your business. If you’d like to learn more, we’re also offering a Virtualization event (see its ad in this newsletter). The time is right for this change, both economically and ecologically.

There’s a lot of buzz about the cloud in the IT industry today. One of the biggest questions I’m often asked is what exactly is the “cloud”? The definition is much simpler than one would think; the cloud is merely the Internet.  The reality is any application or service that is delivered to you via the web is, in essence, a cloud-based service. Pretty simple and not exactly unique! The truth is we have been using the cloud since the inception of the Internet.

I would argue that websites could be considered the first cloud application since without the Internet the sites could not exist. This was soon followed by e-mail and instant messaging services that even the most technophobic people have been using for nearly a decade.

So what’s different today? The biggest thing is that bandwidth has become more readily available, reliable, and cheaper than it was as recently as three years ago. Additionally, many people become more reliant on their cell phone as their primary medium for communication and carry web-enabled devices. These factors have changed the mindset of the application developers who now focus on developing programs that can be accessed via the web from any device that’s connected to the Internet. The expectation of people today is that they can access their data and communicate with others anywhere, anytime. 

So how does this translate to business use? Companies have been leveraging the cloud for quite some time, utilizing applications like Microsoft’s Terminal Server or Citrix XenApp (formerly Presentation) server. These allow companies to leverage the web for remote access to company information, no VPN required.

Another great example of a cloud delivered application is Salesforce.com. This application provides clients with customer relationship management (CRM) services via a public web address and secure login. This application stores and makes accessible a company’s most valuable asset, their customer data, yet no part of the solution resides at the client site. Salesforce.com’s success and track record has helped gain customers confidence in embracing cloud delivered applications.

At Fandotech one service we deliver via the cloud is our Boomerang Offsite backup service. Boomerang automates the backup at our clients’ sites, secures all data for transmission to our data center, and eliminates the need for clients to take tapes offsite.

So is Fandotech a “cloud” company? The definition above allows me to safely say Fandotech has been a “cloud” company for years. We’ve been providing hosted websites, e-mail services, and data center infrastructure for nearly a decade. Newer services have been introduced over the last year (Boomerang Recovery Solutions, 360ITv) responding to customers’ needs for business continuity, disaster recovery, and outsourced infrastructure. Also, customers calling into our technical assistance center (TAC) have received support delivered via the cloud as our technician’s reach into your network environment from afar to assist.

This is an exciting time for businesses as the cloud offers potential to reduce costs for hardware, applications, and support. And you can confidently move forward knowing that Fandotech has been a long-term player in cloud delivered services and is continuing to broaden our scope of services provided via the Internet. 

Brian Doyle

On May 14th we held our first seminar on virtualization and how it can be implemented as a vehicle for savings. This event was well attended, and for those that did make it out, let me first say “Thank You!” For those of you who did not get the opportunity to attend let me give you a brief recap of the event.

The Scouting Report; Virtualization Stats and Overview: Frank Gesino, our VP of Professional Services, led off the event with an introduction to server virtualization. This discussion outlined what server virtualization is and how it is delivered to a client. Frank outlined designing a virtual environment and showed the scale of server consolidation a company can benefit from (often a 10:1 ratio) in this architecture.

Frank also highlighted that savings come, not only in the reduction of servers needed to support your company, but also through increased productivity and reduced downtime achieved by utilizing high availability capabilities found in properly configured virtual environments.

Spring Training; Real-time Virtualization Demo: Joe Bucceri, our CTO, and Mike Conigliaro, our Virtualization Specialist, gave a live demo of virtualization in action. This was the highlight of the event as the duo demonstrated how a virtual Windows server can be built for use on your network in less than 20 minutes!

They also showed the power of the higher availability that can be achieved in this architecture. Mike was working with standard applications on a Citrix server that was part of the VM environment. At the same time, Joe moved his Citrix server from one physical box to another, without any interruption to Mike’s user session.

Regular Season; Integrating Servers and Storage: Bryan Canfield, a Systems Engineer for NetApp, our primary storage partner, presented on the relationship between storage and virtualization. Bryan went over how server virtualization only provides you one component of the high availability architecture and how virtualization of the data store completes the solution. He explained that the image of the virtual server is stored on the server and available for any server in the pool to run, thus allowing that image to be moved in the event of a hardware failure.

Leveraging de-duplication strategies to reduce the overall cost of storage was another highlighted benefit. NetApp’s storage servers have the ability to read data at the block level, and if the server sees that multiple copies of the same file are being stored, it retains only a single copy of that file and makes it available for all necessary services or users. Think of that PowerPoint document that has been emailed to 5-10 employees at 10 MB a person, which is now saved as a single document that can be accessed by all those users; saving the storage for use by other users. This allows data to grow at a more manageable pace, reducing both support and hardware costs.

For those that attended, you will also remember that NetApp can support all virtualization (Microsoft Hyper-V, VMware, and Citrix Xenserver) platforms.

MVP Performance Review; Saving with Virtualization: John Boyd, President of Fandotech, closed out the event, demonstrating how virtualization can save companies cash. In his presentation, John was able to demonstrate that a company with 6 servers today and 50% growth expectancy over the next 36 months would save in excess of $12,000 in power and cooling costs alone and $28,000 in server avoidance costs (cost of hardware, installation, and support).

These numbers illustrate that virtualization is not just for the big companies with hundreds of servers, but that all companies with multiple servers could reduce costs with this technology.

Virtualization can also be a component of a company’s “green” strategy. By moving to a virtualized environment the sample company illustrated above can annually provide CO2 emissions savings equal to three autos, two homes, or the preservation of forty trees.

Contact an Account Manager today for an assessment of how virtualization can help your organization!

Brian Doyle

While working on a datacenter virtualization strategy, I’ve often asked myself if it would be better to migrate a particular physical machine using an automated P2V tool or to rebuild the entire machine from scratch. On the surface, the use of a P2V tool seems like a no-brainer. However, the more I’ve thought about it, the more I could see where a rebuild might actually be preferable in some situations, even if there is a potential for more work.

Advantages of Automated P2V Tools

• Less effort (duh!)

Disadvantages of Automated P2V Tools

• Potential for additional software cost
• Your hardware or operating system may not be supported
• Potential for downtime while the migration is taking place
• P2V software is not infallible. Sometimes it just doesn’t work, and you end up with an non-working VM.

Advantages of Manual P2V

• It’s a chance to refresh your base OS and applications without impacting the production environment
• Potential for less downtime, because your new VM can be built and tested while the existing physical machine is still in production. Moving the VM into production may be as simple as updating a DNS record or a firewall rule.

Disadvantages of Manual P2V

• Potential for additional human cost (e.g. the time to rebuild the machine from scratch)

Clearly, the decision to migrate or rebuild might be forced by the disadvantages of each. In our case, one major roadblock has been older Linux distributions that are not supported by any of our automated P2V tools. This forced us to come up with a strategy for making the manual P2V process as painless as possible. Fortunately, tools like Kickstart and Puppet make it stupidly easy to get a new Linux machine built and configured in a matter of minutes (maybe a subject for a future blog post?), and once the base OS is set up, you’re just an rsync or two away from having a perfect clone of your old physical machine.

It’s interesting to note that we did not always have this streamlined process for building and configuring new Linux machines. Maybe this unintended consequence should be listed as another potential advantage to manual P2V? =)

Mike Conigliaro

Virtualization; it seems that I can’t get through a day without talking about it. It’s definitely the wave of the future, when someone discusses cutting cost, or ease of administration or planning for the future of their network, invariably we’ll bring up virtualization…unless they do first. It appeals to network administrators because administering a couple server platforms is easier than 10 servers. It appeals to management because it’s “green,” yeah, I said you can’t get through a day without hearing about that either these days…thanks Al! The appeal is obvious; ease of administration, lessening your power consumption, cooling requirements, doing more with less, and we have yet to touch on the DR capabilities of virtual solutions.

This is a good chance to discuss the benefits that I’ve seen from down here in the trenches. The number one benefit that network administrators will appreciate, and that I’ve already benefited from comes to light best for remotely managed clients with a virtual environment. It becomes much easier to worry about network connectivity for a couple servers and the SAN, versus the dozen physical servers that we came from. Without getting especially product-centric I will mention that my experiences to date have been with Citrix’s virtualization product, XenServer. The underlying technology does not differ all that much from platform to platform. I’m sure there are some out there that will disagree strongly with that point, but we’ll save that for another post. The management piece for Xen works well, can be loaded on any network node and allows for management of all your servers in one location. This is powerful. The other benefits that I mentioned earlier are the obvious ones that salesmen love to sell and management loves to hear: power, cooling, DR, etc.

DR definitely warrants a mention. Virtualization with shared storage, where your key servers become truly portable. The capability is inherent to allow live synchronization between shared storage from your live site to your DR site, and in the event of a disaster to simply spin up your virtual servers and continue. Very smooth, very powerful!

P2V… Pretty sure there’s been reams written on this topic already, with reams to come. I’ve had good success with Citrix’s XenConvert 1.0, and can’t wait to see their next version.

In my opinion I only see more powerful, complete solutions offered by more software companies on the virtual horizon.

Daniel Kaupp