Closing the Green Storage Gap Page 2
Both economic and ecological mandates can be achieved by balancing performance, availability, capacity and energy (PACE) to meet application service levels and physical floor space constraints along with intelligent power management. Energy economics should be considered as much a strategic resource part of IT data centers as are servers, storage, networks, software and personnel.
Avoiding energy usage by turning off devices when not needed is the intuitive approach to reducing energy usage. For example, turning off or enabling intelligent power management for monitors or desktop servers should be as automatic as turning off overhead lights when not needed. On the other hand, turning off larger and mission-critical servers and associated active storage systems can have a negative effect on application availability and performance. Consequently, avoiding power is not typically a binary on or off solution for most data center environments. Instead, selective power down after analysis of application interdependences and business affect or benefit should be pursued.
From a storage standpoint, the major power draw, based on StorageIO research, for commonly deployed storage systems is spinning hard disk drives (HDDs) and their enclosures, which account for on average 66-75 percent, and controllers account for the balance of electrical power consumption. The evolution from power avoidance is to use power and energy more effectively by implanting solutions with IPM that can vary the amount of power consumed to match required levels of application availability and performance.
Another goal to move toward is IT equipment that is able do more work with the same or less energy and cooling than current technologies. Examples include processors that can do more calculations per second per watt of energy as well as consume fewer total watts than their predecessors across different tiers of servers. Another example would be a storage system that can perform more I/O operations per second (IOPS) or transactions per second or move more data (bandwidth) per unit of energy consumed while at the same time drawing less total power than previous solutions, with different tiers of storage devices. For networks, the example would be to move more data (bandwidth) per unit of energy while the total networking switch draws less power than previous examples.
A Call to Action
The IT industry is shifting from the first wave of awareness and green hype to the second wave of delivering and adopting more efficient and effective solutions.
However, as parts of the industry shift toward closing the green gap, stragglers and latecomers will continue to message and play to the first wave themes, resulting in some disconnect for the foreseeable future. Meanwhile, a third wave addressing future and emerging technologies will continue to evolve, adding to the confusion of what can be done today compared to what can be done in the future.
Items and tasks that you can take action on today to address PCFE and green issues include the following:
- Do more work while using the same or less amount of power
- Leverage faster processors and controllers that use the same or less power
- Consolidate slower storage or servers to a faster, more energy-efficient solution
- Use faster disk drives with capacity boost that draw less power
- Upgrade to newer, faster, denser, more energy-efficient technologies
- Look beyond capacity utilization and keep response time and availability in mind
- Leverage intelligent and adaptive power management modes where practical
- Manage data both locally and remotely, and gain control and insight before moving problems
- Reduce data footprint to enable higher densities of stored data
Greg Schulz is founder and senior analyst of the StorageIO group and author of "Resilient Storage Networks" (Elsevier).