From Meta, a Wikimedia project coordination wiki
This page is part of the Proceedings of Wikimania 2005, Frankfurt, Germany.




Wikidown, wikiup and wikifuture: running the hyper wiki[edit]

  • Author(s): Domas Mituzas
  • License: {{{license}}}
  • Language: English
  • Slides: {{{slides}}}
  • Video: 19:04
  • Note: {{{type}}}

About the author(s): Domas Mituzas is Hardware Officer of the Wikimedia Foundation and one of the people doing the daily server maintenance.

Abstract: Running Wikimedia sites or any other hyper wiki is quite a difficult task. Really.

First of all you've got to be a systems administrator. Having lots of users, means having lots of servers to handle the load. In Wikimedia farm at the time of writing there were 70 servers. There will be more. We have ideas how to handle that.

On the other hand, you are forced to be data administrator. Wiki means having highly dynamic data store for everything - text, images. You have to know your data. You have to improve. We do!

You have lots of users, wanting features. Most of them will be quite difficult to implement on a large website, so history of such events is quite interesting and educating. Also, here pending feature requests should be explained.

And most important - with proper hyperwiki, you have to handle critical conditions. That will be security incidents, power failures, overloads, chain reactions, everything, what is inside wiki administrator's nightmare.

Everyone can join us. Or start their own wiki, to feel that. We're eager to share our pain, joys and experience ;-)