Continuing from last week’s blog with my education from the very erudite Bruce Taylor of Uptime Institute, every data center needs to look at the following five issues for both short and long-term energy savings. These are:
1- Server consolidation, configuration, virtualization
2- Enabling “power-save” features on servers
3- Turning off “dead” servers (no longer in use but running)
4- Pruning “bloatware”
5- Improving the Site Infrastructure Energy Efficiency Ration
Typically, data centers can improve energy savings by 25% to 50% over a two-year period with just by tackling each of these in a cross-discipline way.The Insitute has develop a multi-functional team methodology known as ICE (for Intergrated Critical Environment) to provide both the business and technical rigor required. Not on this list above but a tried and true method is the email sent to all users announcing server shutdown for maintenance for a 24 hour period. Active server users rebel instantly with reasons why their servers can’t be shut down. If no angry responses result, shut them off indefinitely (35 days? More than an accounting period…) and then disconnect the servers. This frees up space, energy, and manpower.
Bruce was very clear that mitigating business risks are as important as energy savings in re-examining, retrofitting, and re-designing data centers. They are at the nerve center of every company, large or small. So active participation and support are required from five key individuals across the organization: representatives from the offices of the CFO, CIO, and Real Estate and Facilities, as well as data center IT and facilities managers and technical teams.
Overwhelmed by the task? Outsourcing of non-strategic activities may be looking attractive. Have a green day!