Google Mojo: The Real Deal in Data Efficiency

December 10, 2009

i know you are up to speed on the Google wizardette Monika H. Henzinger. You know she is pretty good at abstract thinking applied to data issues in massively parallel, distributed computational environments. You may have heard me mention that a certain Google wizardette keeps a low profile and has been know to enjoy the forest. At the Google, according to my research, there is an old fashioned nerd hierarchy. The folks who stick stuff together to make Google Groups talk to Google Apps are smart, but then there are the couple of hundred people at the top of the tech totem pole. Up there with Messrs Brin and Page, Jeff Dean, and 198 others is Dr. Henzinger. She and the Google were awarded a patent for Loadbalancing Multiple Files across Computing Devices.” You can read this document US7,631,310 (granted on December 8, 2009) on Google’s own patent service (sometime soon) or brave the USPTO system. How clever is this invention, filed in September 2004? Read the abstract:

A load balancer evenly distributes processing loads to multiple computing devices. A data structure may be divided into multiple files, each of which corresponds to an estimated load value. The files are assigned to the computing devices in such a way that the processing load at each of the computing devices and the number of files assigned to each of the computing devices is generally balanced.

We have upwards of a million servers, lots of processes, some users, and petascale flows of data. I love that “generally balanced”. This is arguably more complicated than snapping together a couple of Google services. A comparison would be building a kit airplane compared to discovering and producing carbon fiber aircraft components. Mojo, pundits and poobahs, mojo.

Stephen Arnold, December 10, 2009

I wish to disclose to NASA that my reference to carbon fiber is uncompensated. If I even know Dr. Henzinger, I don’t think she would buy me fries with mayonnaise for explaining that she is an algorithmic satrap.

Written by Stephen E. Arnold · Filed Under Google, News, Technology

Comments

Comments are closed.

Search the site
Subscribe to Beyond Search
Feature archive
News archive

Stephen E. Arnold monitors search, content processing, text mining and related topics from his high-tech nerve center in rural Kentucky. He tries to winnow the goose feathers from the giblets. He works with colleagues worldwide to make this Web log useful to those who want to go "beyond search". Contact him at sa [at] arnoldit.com. His Web site with additional information about search is arnoldit.com.