Archived Content
The following content is from an older version of this website, and may not display correctly.Online user-generated encyclopedia Wikipedia was down for more than one hour on Wednesday evening after temperature rose in an Amsterdam data center that hosts the organization's servers, leading many of those servers to automatically shut themselves down to prevent damage.
Switching off of the Amsterdam servers caused access issues for the site's European users, but when Wikipedia engineers enabled a failover procedure to reroute traffic to the company's primary infrastructure in St. Petersburg, Fla., it turned out the failover mechanism was broken, bringing Wikipedia sites down globally, Wikipedia's Mark Bergsma wrote in a blog post. The problem was caused by dysfunctional DNS resolution for the organization's sites.
It was not clear on Thursday how long the outage lasted exactly, but Bergsma's first update to the original notification of downtime said it had lasted for longer than one hour. The final update, notifying users that the company's SSL gateway was back up, came at 21:32 UTC on Mar. 24.
Wikipedia keeps its European IT equipment in data centers operated by EvoSwitch and SARA in Amsterdam.
US servers are installed in Switch and Data and Hostway data centers. The organization has a total of about 350 servers, an infrastructure it is planning to expand after receiving a $2 million grant from Google in February.
Switching off of the Amsterdam servers caused access issues for the site's European users, but when Wikipedia engineers enabled a failover procedure to reroute traffic to the company's primary infrastructure in St. Petersburg, Fla., it turned out the failover mechanism was broken, bringing Wikipedia sites down globally, Wikipedia's Mark Bergsma wrote in a blog post. The problem was caused by dysfunctional DNS resolution for the organization's sites.
It was not clear on Thursday how long the outage lasted exactly, but Bergsma's first update to the original notification of downtime said it had lasted for longer than one hour. The final update, notifying users that the company's SSL gateway was back up, came at 21:32 UTC on Mar. 24.
Wikipedia keeps its European IT equipment in data centers operated by EvoSwitch and SARA in Amsterdam.
US servers are installed in Switch and Data and Hostway data centers. The organization has a total of about 350 servers, an infrastructure it is planning to expand after receiving a $2 million grant from Google in February.