Yahoo patents load balancing, my computer breaks
I ran out this morning, and bought a new computer. The old one died on me yesterday.
It’s a good thing for external hard drives. I would have been pretty upset if I had lost all of the data on the old computer.
I did have to do a lot of updating and installation of software, and it might have been time to get a new computer anyway. I might have to pay more attention to how Yahoo handles their computers.
(Will be back to semi-regular posting as soon as I do a little catching up.)
In the meantime, I had a few minutes to take a look at the US patent office tonight. Yahoo! was granted a patent today on a filing from 1999 on coordinating information between multple servers that share information, and also on servers that may cache some of that information.
Web server content replication
Filed August 19, 1999
Inventors: Stanley Yamane, Swapnil Shah, Keith Weng, Arthur Mateos, Marco Lara, and Jason DeBettencourt
Changes to files in a master source file set on a master computer are identified. Local host addresses of the files are converted to uniform resource locators. The identified changes are stored in a modification list comprising the uniform resource locators specifying changed files of the master source file set. The modification list is transmitted to one or more computers, notifying the one or more computers that the master source file set has changed. A response is received from at least one of the one or more computers, indicating that the identified changes are installed. An alert is issued indicating that a copying of the changes by an agent in at least one of the one or more computers has failed, to cause a traffic manager that redirects traffic to decrease requests made to a web server on the computer having the agent
There’s a huge list of patents cited as references in this one, going back to 1988. The follow-up list of other cited documents is pretty large, too.
If you are interested in topics like web servers, caching, analytics involving search engines, load balancing, and the technical aspects of updates to information stored on a search engine or directory, some of the early part of this patent is fairly readable (or should I say “relatively” as opposed to some of the later parts of the document).
One element that I found interesting is the section on routing around bad information, or updates that don’t go as expected. Yep. Sometimes search engines even have technical problems.