Data Center Automation is more than a bunch of software tools that do the work that normally is done with manual intervention. In fact, automation follows a maturity model in which interdependent stages must exist for success.
The first phase of any automation strategy must be to fully understand what your IT and application environment consists of, even to the most granular level. This is particularly consequential when undertaking a server consolidation project without a complete understanding of critical application dependencies across server platforms as well as their supporting infrastructure components. It is very difficult to provision, or especially de-provision servers and applications without a comprehensive knowledge of the impact of the change, subsequently introducing the risk of a potentially severe interruption to the business. Many have found it much easier to simply keep outdated and decommissioned servers on-line rather than risk implications of taking a critical application off-line, simply because they were not sure of the outcome. It only makes sense to take the guesswork out of server provisioning and de-provisioning, crossing ones fingers hoping for the best is simply not an option.
Having a detailed and granular understanding of your environment, as well as mapped dependencies, provides the foundation for configuration standardization as well as control around change. Setting a gold standard for server configurations eliminates a huge culprit in potential service disruption. Leveraging gold standards to detect configuration drift insures that all servers comply with the best-known configuration standard without risk of variations. This eliminates unauthorized changes that could lead to devastating results. For example, if there is an unauthorized change in a server’s port number to support a new application deployment, this could affect other dependent applications which relies on the original port value to properly operate. Manual detection, or troubleshooting, could result in countless hours of lost productivity and revenue. Having change and configuration standards as well as configuration change detection in place ensures that this type of problem would not have gone unnoticed and can be quckly rectified.
In order to successfully automate, you must first discover. But remember the discovery phase is still one of the many entry vectors into a comprehensive data center automation strategy!