Last week Twitter announced changes to the search functionality on Twitter.com, adding some advanced elements that expanded the results to include user profiles.
Over on the official engineering blog, Twitter’s search engineer Krishna Gade has written about how modifications to Twitter’s search engine (which began in the Spring of 2010) led to the company replacing their Ruby-on-Rails front-end with a new Java server, codenamed Blender.
End result? Twitter search is now three times faster.
Gade uses the recent spike in search queries caused by the Japanese earthquake and tsunami to illustrate the benefits of the new system.
Twitter search is one of the most heavily-trafficked search engines in the world, serving over one billion queries per day. The week before we deployed Blender, the #tsunami in Japan contributed to a significant increase in query load and a related spike in search latencies. Following the launch of Blender, our 95th percentile latencies were reduced by 3x from 800ms to 250ms and CPU load on our front-end servers was cut in half. We now have the capacity to serve 10x the number of requests per machine. This means we can support the same number of requests with fewer servers, reducing our front-end service costs by an order-of-magnitude.
From here, things get a little bit technical, but head on over to the engineering blog for a lot more detail.
In the meantime, we’re curious – does Twitter search seem faster to you? Hit the comments to let us know.
(Source: Twitter engineering blog.)