Everywhere you turn, people are talking about “the cloud”. Most of those people aren’t customers; they are service providers and software companies. Despite a lot of over-hyping (i.e. Microsoft’s “to the cloud!” commercials), I have personally spent time working with a few different cloud platforms and have truly been amazed at what they can do. While there are many different types of cloud computing, when I think of “the cloud”, I think of IT as a Service (ITaaS) so that’s what I’ll cover in this article.
Now cloud computing isn’t just “hosted virtualization”, it’s much more. There are high-end enterprise-grade features that only select enterprises have and medium-size companies and SMBs just can’t afford. Cloud computing is going to bring that level of features to every company and make it affordable. It’s easy to have cynicism and skepticism about cloud computing but, trust me, cloud computing is powerful and I’m excited about its future.
What’s so cool, you say? Here are 10 reasons that cloud computing is cool:
Why should you have to pay for an entire physical server to run a single application? Many times, the total cost of ownership isn’t taken into account. Plus, over time, that one server will cost a lot more in electricity, cooling, and datacenter space than it will cost just to buy it.
The premise behind cloud computing is that you only pay for what you use. In other words, if all you need to do is run a single app that takes few CPU cycles and MB of RAM then that is all you will pay for. With some cloud companies that means that you supply a credit card, have no contract, and get automatically charged a monthly fee that is only tied to how many computing resources that you used. VMware’s vCloud Express is one example of a service like this.
With VMware vSphere, taking snapshots are common practice. Admins use snapshots daily to take a “picture” of the state of a VM at a point in time before changes are made. With those snapshots, they can jump forward and back in time. Even with vSphere and a solid enterprise infrastructure, taking snapshots can take time and you can only take a snapshot of one VM at a time.
With cloud computing, you will be able to take instant snapshots of multiple virtual machines (teams). Even better, you’ll be able to instantly use those teams/groups of virtual machines.
Picture this –you have a working Exchange Server config with four Exchange VMs, a Windows AD/DNS VM, front end web server VM, and simulated end-user client VM. Perhaps you use this team of VMs for training or development on Exchange. With one click, you could instantly duplicate this entire team of VMs, one or more times, and use them individually. You could even setup web access for your students wanting to learn Exchange to access the consoles of these teams of servers.
I’ve done this with the Skytap Cloud and it’s really amazing. Consider how much work and time it would take to create this even with vSphere (the most advanced virtualization platform available today).
So much focus is put on choosing between a private cloud (internal virtual infrastructure) and a public cloud. You should know that there is another option – Hybrid cloud. With hybrid cloud your private cloud (or just a single internal server) is connected via secure VPN to the private cloud so that the two can work together. You could even connect all your users securely to servers in the public cloud.
Most people think that cloud servers can only be connected to the Internet to provide, let’s say, public facing web servers or applications. Hybrid cloud opens up a third, and much more practical, option for companies of all sizes to use cloud computing.
Here’s what VMware’s Hybrid Cloud diagram looks like:
Implementing features like HA and disaster recovery in your existing virtual infrastructure can be expensive time consuming. Because the cost of these features can be spread across, users of “the cloud” features like HA can be affordable for all. Plus, with simplified web interfaces, an advanced feature like HA can be easy to use (or even “automatic”).
Similar to HA, disaster recovery is also an advanced feature that many company struggle to implement and afford. With cloud computing, the service provider assumes the risk, cost, and burden of implementing disaster recovery.
Note: disaster recovery is a sensitive subject and I don’t encourage you to put 100% trust in your service provider. You should always have your own, internal, backup plan if the service provider has a disaster. I would encourage you to talk more with your provider about this topic and find a way to have your own accessible offsite backups in case a disaster hits “your cloud”.
Many companies struggle with taking backups of their servers. The daily monitoring of those backup jobs, changing tapes, and moving those tapes offsite requires a lot of time and effort. Perhaps moving your servers to “the cloud” can eliminate this work for your IT group. Even better, perhaps the cloud provider has offsite disk-based replication such that they don’t have to worry about tapes either, the data is automatically offsite, and the data can be instantly available.
When I was an IT Manager, all our hardware was on a 3 year lease. Because those leases occurred throughout the year, which meant that we were constantly swapping hardware out for newly leased hardware. A single server swap involves spec’ing out a new server, ordering it, unpacking it, “racking” it, cabling it, loading the new OS, and then migrating data. With 100+ server, just keeping up with the constant hardware swaps to keep out equipment current was a large strain on the IT group.
With all servers virtualized in the cloud, the cloud provider has to deal with hardware swaps and virtual machines can be “vMotion’ed” from the old hardware to the new hardware.
Apple’s iTunes is an example of a very successful “software catalog”. With cloud computing, most providers offer a software catalog where you can not only select the software you want (and pay for it, if applicable) but simply import pre-created virtual machines that contain that software into your cloud infrastructure.
Let’s say that you want to create an Apache web server with a MySQL backend. You could just click on those two VMs from the catalog and have them up and running in seconds or minutes. Compare this to what you would do in a typical virtual infrastructure - download Linux, install it in a VM, and load the applications.
Here is what the Skytap software catalog looks like:
Cloud computing wouldn’t be very cool if you couldn’t move your virtual machines in and out of the cloud at will. Someday, this will be quick and easy but, today, with the size of virtual machines (10+ GB) and the size of most people’s bandwidth to the Internet (between 1-10MB/sec), it takes time to transfer a virtual machine between a private cloud and public cloud and it just can’t be done with the VM powered on.
Transferring a VM from your internal virtual infrastructure to “the cloud” is usually done by FTP’ing the virtual machine disk file (VMDK if vSphere) to a cloud host. Then, you create an import job in the cloud to bring that VM in (an export works the same way).
What is amazing is when cloud providers allow you to do all of this as a “self-service” function. That means that you could import or export a VM, all yourself, with compatibility between the private and public cloud virtual machines.
One of the big selling points of the cloud is that you can move from a capex to opex expense model. What this means is that instead of you having to create a capital expenditure (capex) request (and get it approved) every time you buy a new server (which is a HUGE pain at most companies), with the cloud you move to a monthly (operational expenditure / opex), pay as you go model (similar to moving from “buying” a house to “renting” a house). The CFO & financial people at your company will usually like this model better as they don’t have to lay out a chunk of cash one month for large server expenditure and then depreciate it over time.
While cloud computing, in its current form, may not be the perfect fit for every company, the goal of cloud computing is to save your company time and money (and that is a perfect fit for every company). Your time can be freed up to do whatever it is that your company does best and let a cloud provider worry about things like server swaps, datacenter maintenance, capital expenditure, backups, high availability, and disaster recovery. Those aren’t things that make your company money – just expensive & never-ending IT maintenance tasks.
I’m not saying that EVERY cloud company offers the features I discussed in this article or even that you can find all these features from a single company. Every cloud service is different so make sure you spend some time educating yourself and shopping around.
In summary, cloud computing has a lot to offer you and it deserves to be investigated. More next features are released every month, making it more and more compelling. There are many ways that cloud computing saves you time and money and I recommend to all admins that they start their research and education process now to see if “the cloud” is a good fit for their company.
You may not know it, but System Center Virtual Machine Manager can be used for…
As we adjust to a new remote work culture due to coronavirus, a secure VPN…
Every Exchange admin lives with the constant fear their system will be breached. Having SPF,…
A GE data breach exposed a hacker’s treasure trove of employee records, including Social Security…
Ready to go back to the future? Here’s a look at some Linux text editors…