IEEE Spectrum

A More Efficient Way to Process Tasks in the Cloud

You only needed one tomato. Unfortunately, the three people in line ahead of you at the supermarket have loaded their carts like they’re on an episode of “Doomsday Preppers.” You scan the other checkout lanes, but even the express line is backed up because some dude insists on paying in spare change.

A similar problem bedevils the cloud services industry. Google Maps, for example, must parse simple tasks—for instance, directions to a friend’s house—without getting bogged down by those more complicated—go to a friend’s house, but avoid highways and add a stop at the supermarket to pick up some tomatoes. It, and other cloud-based services, accomplish this thanks to tools that send incoming tasks to different servers, effectively forming queues according to size and complexity. 

Now, the developers of a new task scheduling tool say their program beats the competition at managing and prioritizing such tasks in an efficient way.

By admin

I'm awesome! What else would I say about myself.