This PowerUp blog entry was written by Andrew Hillier, CTO and co-founder of CiRBA. In that role, he acts as the visionary for product strategy and CiRBA’s technology roadmap. He has more than 20 years of experience in the creation of mission-critical software for some of the world’s largest financial institutions and utilities. His articles have been published in leading IT and online publications.
The combination of PowerVM and AIX provides a rich foundation for hosting mission-critical workloads. But their flexibility and advanced nature can also create significant planning and management complexity. Minimizing operational risk while driving efficiency can be challenging. Here are five ways to maximize service levels while optimizing infrastructure:
- Define operational policies that map LPARs (and the applications they support) to the underlying infrastructure. How big LPARs should be, their entitlements and their placements on hosts (and within shared pools) can significantly impact how they operate. Don’t overprovision in the name of safety, because this can greatly impact overall efficiency. It is better to scientifically control the safety margins than to make them arbitrarily large.
- Get a proper handle on the “pipeline” of new applications and upcoming workload demands on an environment to forecast capacity requirements. Modeling the impact of upcoming “bookings” enables more accurate capacity management. The alternative is building far more capacity than required, just to deal with the unknown.
- Define separate, detailed operational policies for virtual I/O servers. These are critical to proper system functioning and application performance. Clear rules that govern how big they are and how they should be resized are critical.
- The same is true of shared resource pools, which are increasingly used to control licensing and limit resource utilization. How big they should be and how they should be resized if they are too full (or too empty) should be clearly defined and controlled.
- Do not do all of this by hand. Spreadsheets and sticky notes are no longer sufficient–modern AIX systems require advanced analytics, driven by policy and diverse data feeds, to manage all of the moving parts.
The shift toward analytics is the natural evolution of management in these complex environments and can have a significant ROI. PowerVM and AIX are the enablers of efficiency, but this efficiency can only be realized when you are clever about how these technologies are used.