Are You Driving Virtualization With Eyes Wide Shut?

As virtualized environments grow larger and more complex in design and architecture, it's imperative to be able to spot current and potential bottlenecks that could lead to performance degradation and outages. Without a bird's-eye view of your environment, you may as well be operating with your eyes closed. There is no way to guarantee you have a reliable environment if you don't know its health. It can be difficult to get total visibility into the health of a virtualized environment with just a single product. However, having a few different solutions for monitoring the health of your environment is not necessarily a bad thing. And although the there is additional cost, consider the far greater cost associated with an outage. I guarantee most monitoring and event management systems pay for themselves once they help you avoid an outage that could otherwise have cost thousands of dollars. Being able to quickly find the root cause of an outage is an important win, since some outages can't be avoided. The more complete the view into your environment, the better off you will be. The list below includes some key areas you should have a quick way of monitoring:

• Network
• Storage
• Server/blade
• Hypervisor
• Application

  Different parts of IT may manage each of these areas individually; so the next question would be, "Can you correlate the information you gather in multiple systems into a single view?" On countless occasions, I've had to troubleshoot a problem without having the complete picture. This can make it impossible to truly know what the root cause of an outage was, and means wasted time searching for a silver bullet that can never be found. We've all heard complaints about slow-running applications: But what does that really mean, and where do you go looking first? VMware's vCenter Operations can provide comprehensive visibility into your constantly changing vSphere environment. You'll get visibility into your infrastructure and applications with at-a-glance dashboard views backed by in-depth analytics. You'll quickly solve problems and be able to take a proactive approach, with the ability to spot trends and create performance baselines. Another product in this space that I like is vFoglight from Quest. It offers a lot of the same high level and detailed information that can be used to see and understand more of your environment. Also, don't forget to take advantage of monitoring and alerting tools that may come out-of-the-box with certain products. Tintri provides a great deal of information out-of-the-box in an easy-to-read format. This makes troubleshooting performance bottlenecks faster, and allows analysis of the storage layer directly from vCenter. Tintri uses the same approach to overseeing the environment down to the storage layer, with simple high-level metrics and the ability to drill down into the details. Whichever product you use, the key is to make sure you can get complete coverage, or support for pulling in information from other monitoring and event systems. The ultimate goal is to be able to make the information useful, even at a high level. When there's an indicator that something is wrong, IT should be able to hone in and fix the problem before the customer calls.