Patching with a smile
Managing images is hard. We have all been there. The processes of adding new versions of applications is cumbersome while the amount of changes is massive. The complexity of all the different components in your image also interfere with each other, both performance and functional wise. So how do you know by the end that everything still works: functionally, technically and in performance?
And how many of us still use a manual process to manage all these different images? Do you ever test the coherence of the image and the different applications in the image after each update or patch installation?
How often do we dare to patch or upgrade an image, given the complexity and the uncertainty of the outcome? How many of us wait before we upgrade until we have a big pile of changes to implement? This manual process is so time consuming that we do not do it often, and when we do it we take shortcuts when testing or validating the outcome.
What happens if something goes wrong with that large amount of changes: Which change caused the issue? Was it the first patch you installed? Or was it the one from Microsoft or was it patch 87 that updated an application that was ‘home made’ developed? The longer you wait to test or validate, the more complex the problem becomes.
What more do I should say… you must test and automate the changes in your images and get it right first time. How, you ask? By automating the process of applying and validating the changes you have made continuously before you take anything into production. Easy to say, you might think, but it can be done! There are many tools out there that help you automate the creation and management of images, known as ITSM software solutions.
Validating those changes and seeing the impact on performance, and the interference with other applications, is a process that needs some deliberation. Of course you can keep on using error prone manual processes, alternatively you can automate that process and avoid taking the risk.
A virtual user could be the perfect solution to this problem. It allows you to install a smaller number of patches at a faster pace, validating them at the same time—avoiding high risk in the process. A virtual user measures the impact of any change on performance and validates if an application (or applications) still works before you take anything in production. It behaves, after all, like a real user.
Automating the process speeds up the implementation of change. Using a virtual user guarantees the validation of those changes in an existing environment. Faster implementation of patches means smaller amount of changes at the same time, less risk, easier to manage, and certainly easier to find potential issues when using a virtual user.
Don’t be afraid of change, quit your comfort zone and patch with a smile.
Jasper Geelen & Jos Dikhoff