The Autonomic Element of ODC
One really clever thing about ODC is the way it builds on existing systems. No need for forklift upgrades! This is an interesting aspect of much successful modern technology, where some preexisting system is used as the basis for the next generation—mobile phone text messaging, Java on mobile devices, broadband Internet access over legacy copper cable, TV set-top boxes, IP/MPLS on telecom networks, and so on.
The merits of this approach are backward compatibility (for example, you can still use the voice service on your mobile phone, along with a growing range of other services); low cost of service access (you don't necessarily have to buy a new mobile phone to use the new services); and reduced risk for service providers. ODC employs a similar model by defining five autonomic transformation levels for the full migration:
- The basic or manual level is pretty much where we are today: multiple disparate sources of management data (non-integrated proprietary consoles, log files) watched over by often large numbers of highly skilled staff. In the basic autonomic level, staff often initiate data collection in advance of analysis and reconfiguration.
- Managed-level organizations employ (management) tools to collect and consolidate data. IT staff then carry out analysis prior to tweaking operational systems. While the staff are still very skilled, they can operate on a somewhat higher abstraction level than their peers in the basic level.
- At the predictive level, operational systems are beginning to be seen as a cohesive entity. This abstraction is facilitated by the use of advanced collection and analysis tools. These tools correlate data and perform such tasks as root-cause analysis. The predictive level significantly differs from the basic or managed levels in its use of action recommendations. IT staff are presented with specific actions to remedy problems. Clearly, as we move up the autonomic levels, the grunt work is getting done by software rather than humans. This reduces the dependency on arcane IT skills and improves decision-making and response times.
- In the adaptive level, the human input is reduced still further as the system takes action to remedy problems. IT staff can manage the performance of operational systems against defined service-level agreements (SLAs). In cases where IT has been outsourced, this emphasis on SLAs can be very welcome in reducing costs; for instance, if the host organization sets financial penalty limits on downtime.
- IT staff focus on facilitating the business needs in the autonomic level. Software is employed to dynamically manage the operational systems based on specific business policies. The autonomic level integrates IT into the core business.