Sometimes Scott Adams is so right it’s easy to see his developer roots. Yesterday’s Dilbert was spectacular:
Maybe this hit home because we actually do virtualize apps, and as a logical consequence it was actually pretty easy to implement our EnergySaver feature. In practice, if an app’s existence is spread across some number of server or vm instances and is therefore NOT dependent upon any one particular server or vm instance, then it’s easy to automatically turn on / turn off computers as needed without impacting the app itself.
That’s what we did with our technology preview for EnergySaver, which works great with any fabric app (no further work by the developer needed – purely an operational policy).
Datacenters are increasingly feeling the pressure to reduce energy consumption, particularly as they “feel the need to scale”. As a result, this ability to easily drive down energy consumption is pretty handy … especially since it actually is so easy to do in a deployed fabric.
One more note on this Dilbert. I think we should define a unit of time called something like the “interframe-moment”, which is the average amount of time that someone who just doesn’t understand how to do something expects it’ll actually take to build that same something. That was just great!
Comments on this entry are closed.