Skip to main content

Storage Power Modeling

In order to include energy considerations in storage management, we attempted to identify the storage power consumption components, model their behavior, and understand the trade-offs between storage performance and power. Towards this goal, we developed a model for the power consumption of the different storage controller components for various workloads characteristics.

The majority of the storage controller power consumption is attributed to the disk arrays. Of the power attributed to the disks, about two-thirds of the power required is used for spinning the platters; these platters are spinning even when the disk is "idle". The other third of the power is used to serve I/O requests. In idle mode, the disk is spinning at its full speed, but no disk activity is taking place. Staying in idle mode when there is no disk request provides the best possible access time since the disk can immediately service requests. But it is also the least efficient use of power.

To estimate the power consumed by a given host workload, our method translates the workload to the primitive activities induced on the disks. In addition, we identified that I/O queues have a fundamental influence on the power consumption.

The energy consumption of the storage server depends on the aggregation of various application workloads. This makes it difficult to proportionally divide the energy cost of the storage server among the different applications. In our model we assign storage energy consumption per application based on the application's file usage and storage configuration.

The storage server energy consumption is composed of fixed and dynamic portions. The fixed portion is independent of the workload, while the dynamic portion is directly affected by the application workload. The application workload is translated to values representing the quantities of primitive storage activities, such as the number of disk seeks. This is done using performance statistics provided by the storage server. Most storage servers provide statistics on the caching and workload type for each array and logical volume. The energy consumption of each physical component is computed using interpolation and the energy consumption dataset. Our logical volume estimation framework deals separately with fixed and dynamic energy consumption.

We integrated our modeling into a power-aware capacity planning tool (DiskMagic) to predict system power requirements. We then integrated it into an online storage system to provide online estimation for the power consumed. DiskMagic is a capacity planning tool that helps build a required configuration to support given workloads. The power models integrated into DiskMagic provide customers with details of what will be their energy consumption running these workloads.

The storage energy estimation framework is now freely available from IBM's Open Process Automation Library (OPAL) . The framework is implemented as an add-on module for IBM Tivoli Monitoring (ITM) and Tivoli Storage Productivity Center (TPC) customers.