The underpinnings of IT automation

Last week, Mary Anne Winniford wrote about IT process automation as a mixture of sub-markets that, as a group, are helping to move some IT organizations towards superior levels of operational efficiency and accountability. The submarkets she mentioned include service desk related technologies such as workflow, data center rooted technologies, such as job scheduling, as well as more whole-cloth forms of IT process automation such as top-down and bottoms-up process mappings. These support well-defined processes and business alignment (in the case of the former) and a software-centric approach to something that might be called “run-book automation” (in the case of bottoms-up approaches).

In this column I’d like to step back and suggest that IT process automation is really an end-game in which process and best practice initiatives must come together with software-centric approaches to task automation. Since the industry is far from ready with adequate and complete technologies for doing this pervasively, IT is well advised to focus on preparing for IT process automation first, and implementing it second, on an opportunistic, when-appropriate and when-possible basis. I believe that IT process automation is likely to be a huge area of growth for the IT management industry in the coming 5 to 10 years, but that it will only really fly once core architectural underpinnings are in place.

Today I’d like to examine just a few of these underpinnings:

* Discovery. Actually, the key word here is “visibility.’ Automating processes via management software can’t be done if there isn’t an accurate and current version of IT reality to work with. OK, call it a “CMDB” if you like, but it’s really not limited only to CMDB implementations per se. Discovery can range from Layers 2 and 3 network-centric discovery, to application-centric ecosystem mapping, to discovering traffic flows and usage.

* Identity management and access control. Don’t tell me that you want to implement process automation without good tools to manage user access and entitlements across the infrastructure – both within and beyond IT professionals. IT process automation will depend on clear and accurate information about who is present, as well as who has done what to the infrastructure, and when. Having these capabilities will also extend to support areas such as compliance and IT governance.

* Configuration management capabilities are also critical for IT organizations seeking to implement a more automated approach to managing change. These will also support visibility requirements while providing active areas of automation. As such they are both central to IT process automation, and potentially low-hanging fruit when the right technology investments are in place.

* Analytics. I’ve written about analytics in the past and it will become a big part of EMA’s research this quarter. Analytics range from upscale event filtering to fuzzy logic and chaos theory with OLAP somewhere in between. Analytics are central to virtually all IT disciplines. Root cause diagnostics, chargeback accounting, effective and dynamic capacity optimization are just a few examples. Without good analytic investments, a heavily workflow-dependent approach to IT process automation will be like sticking together horses and buggies and a relay race against, well, Maseratis.

* Ultimately even a CMDB may become a critical resource for IT process automation as it can provide, in theory least, a trusted and current view of infrastructure components as they may to specific service requirements. CMDBs can be good and successful initiatives if they’re well focused and designed to roll out over time. In the coming years, they may in fact become required cornerstones for effective IT automation, as technologies and standards to support CMDB come of age and IT experience with CMDB matures.