Hello everyone.
Just got back from VMware’s Virtualization Forum on October 8, 2009 in Orlando and wanted to share some thoughts:
If I had to sum up the theme in one word, it would be “cloud”. Cloud Computing is the latest buzz word/hot topic to hit the industry, following in the illustrious footsteps of Software as a Service (SaaS), Consolidation, ITIL, and Green IT. The term cloud computing purposefully covers many topics, but one of the keynote speakers at the forum representing Enterprise Strategy Group summed it up very nicely: Cloud computing is a means of consuming and distributing computing resources. Still confused? Let’s take a closer look.
Most companies utilize Microsoft Exchange Server as their Email Messaging solution. Ten years ago, if your company needed an Exchange Server deployment, the organization would have purchased a new server according to certain specifications, waited a week for its arrival, installed the operating system and patches, and finally installed the Exchange Server component, which would allow users to send and receive corporate emails via the Microsoft Outlook client. If the company’s email use or employee count grew unexpectedly and the server suddenly ran out of disk space, required a RAM upgrade, or if the motherboard went bad, new hardware was procured in order to remediate the issue. The preceding scenario (which could just as easily apply to your Line-of-Business application or your File Server) inevitably resulted in poor service to users and customers and days of lost productivity.
If the same situation occurred 3 years ago and your Exchange Server had been installed on VMware’s VI3 Platform, the best case scenario is that the Server could be shutdown, the virtual RAM added or the virtual disk space expanded, and then be brought back online in only a matter of minutes. If the Physical Hosts’ motherboard went bad, the Exchange Virtual Machine would have rebooted on another Host in the VI3 Cluster and little productivity was lost. In the worst case scenario (an infamous Windows Blue Screen of Death), a new Virtual Machine could have been deployed and Exchange data restored from backup in under 2 hours.
If this same scenario happens today on VMware’s vSphere platform (whether you manage this in your network or pay as you go from a “Cloud” Provider), the RAM can be added or the disk expanded on the fly with zero down time. If the motherboard of a Host goes bad, the Virtual Machine is not disrupted at all through vSphere’s Fault Tolerance feature. If you receive a Black Screen of Death, well, you’re still in bad shape, but the advancements in backup technology, specifically Virtual Machine-aware backups, result in much lower time to restore.
An illustration of how an internal and private cloud integrates with an external and public cloud.
What will these scenarios look like 5 years from now? What if the applications were ‘smart’ enough to request more RAM or more disk space when needed, and what if whatever platform that application was running on was smart enough to grant this request on the fly? Lofty aspirations, to be sure, but it’s coming, and it’s the essence of Cloud Computing. And what if that new Exchange Server didn’t need Windows anymore? What if it was available as a pre-packaged Virtual Appliance custom built to your company’s requirements that could be downloaded and operational in minutes? Well, that would take care of the Blue Screen of Death.
