A Practical Guide to VDI Change Management
Part 6: Handling Change in VDI Production
Monitoring solutions are good for fixing problems, but what is the best way to avoid problems and how do you learn to anticipate and prevent them completely? In this article, we find answers to this question as we continue our 8-part blog-series on a practical guide to VDI change management.
In a healthy VDI production environment, the main enemy is change
In chapter 2 we concluded that the chance of failure in VDI is relatively high due to infrastructure complexity. We also concluded that a failure in VDI carries a heavy negative impact on organizations due to centralization. In chapter 3 and chapter 4 we see that the amount and speed of changes that need to be digested is growing fast.
The combination of these factors makes it critical to implement a comprehensive change impact process that covers every type of change (see the below picture) that may hurt your IT operations, and as a result may damage the organization’s Business as Usual (BAU).
|Chance of Occurence||
Impact on Organization
Impact on Organization
|Facts of Life||
- Table: Different types of change can disturb IT operations and therefore business continuity
The changes discussed in chapter 3 and chapter 4 are examples of planned changes. These types of changes are known, anticipated, and explicitly implemented and processed by the responsible IT staff. Changes of this type include new builds of Windows 10, new versions of business applications (like SAP or Epic), regular weekly or monthly updates (e.g. “Patch Tuesday”) and all significant security patches. The impact on performance and density of systems can vary in the range of 0% to 30% (L1TF) to 50% (EHR vendor) but the impact can be predicted and therefore be anticipated (if properly tested!!!).
More and more applications and infrastructure software are automatically updated (examples include AWS and security software). Most of these updates will not have a performance impact but some may. This type of automatically processed change falls in the category of unplanned changes. Unplanned changes may also be caused by someone else in the organization that altered a switch or added or deleted something.
Next to the planned and unplanned changes the regular daily use of IT environments can also cause a slowdown of desktop and application performance called gradual deterioration, which may be caused by the pollution or saturation of registries, storage and/or other devices.
The last type of change we need to plan for are those rare occasions that come with a very big impact such as disasters or other major incidents that can be caused by nature or have human causes.
“The only constant in life is change” - Heraclitus of Ephesus (535 BC – 475 BC)
Why traditional stack monitoring solutions are not enough
Traditional stack monitoring, or End User Monitoring (EUM), solutions are infrastructure focused by nature and will tell you all about CPU, I/O and memory health. With their strength in measuring and analyzing real user activity they are reactive by nature, offering effective solutions for root cause analysis. This way they help organizations locate and fix IT problems that occur during production.
Some vendors in the End User Monitoring space also use predictions based on the extrapolation of historical data. In environments where professional testing solutions are not available this can add value but this is certainly insufficient when a comprehensive change management process is needed, such as in health care, finance and other verticals where the cost of failure is high and very tangible.
- Picture: Stack monitoring solutions are effective for problem fixing but not for problem prevention
Some EUM vendors recently introduced light versions of the virtual user concept to add a pro-active concept to their monitoring solutions. These solutions do start applications to test availability but lack real-life simulation of in-depth application user experience as is needed for key business applications.
Modern organizations looking for effective ways to safeguard business continuity (often after serious outages) must complement their infrastructure-oriented monitoring solutions with 100% pro-active, strict end-user experience perspective focused, test solutions, designed to 100% prevent problems.
To safeguard a delicate, complex, and business critical centralized Windows environment, prudent organizations must leverage a comprehensive pro-active testing environment to handle and prevent possible negative impacts of all the four types of change as described above. Combining best of breed monitoring solutions with a best of breed testing solution will significantly reduce business outages.
Conclusion: Monitoring solutions are good for fixing problems, testing is best for preventing problems.
Did you miss the previous article or want to read them again? Simply click the below links to review our Practical Guide to VDI Change Management in full:
Read Part 1: IT Change Management in General
Read Part 2: Why VDI is Very Sensitive to Change
Read Part 3: Change Accelerates with Windows 10
Read Part 4: IT Change Accelerates in General
Read Part 5: Building Resilient VDI Infrastructures
Read Part 7: Change Impact Management for VDI
Start using Login VSI today
Our industry-standard software is built to help you avoid problems, lower costs and improve performance. Request a Quote or get your (free!) Trial below, and benefit from our award-winning services.
About the company Login VSI
The company Login VSI provides end-user performance insights for virtualized desktop and server based computing environments. Enterprise IT departments use flagship product Login VSI (for load testing) and new addition Login PI (for continuity testing) in all phases of their virtual desktop deployment—from planning to deployment to change management—to build and safeguard a good performance, a high availability, and (as a result) a good and consistent end-user experience. For more information about Login VSI or for a free test license contact us.