2 comments

[ 2.6 ms ] story [ 11.9 ms ] thread
Great infographic and overview.

I'm interested in why such low CPU usage is a good thing on the app servers. I understand the need for handling spikes, but they're saying in the infographic that max is 12%.

My uninformed, initial thinking was "That's a lot of wasted CPU" on the app machines. Can anyone explain why spending the extra money and having 88%-95% unused CPU vs. scaling back the number of machines and running them at 50% load?

Could be the number of concurrent requests they can serve by scaling out to more servers