Latency: from a dream to nightmare in 100ms

"It’s the network!” is a reflex of every software engineer, and they are very often right. The Internet is much more alive than we think. But when our applications are online how can we work with unpredictable latency? Can we optimise it?

Since operating Algolia's servers in 38 different datacenters around the world, we came across various head-scratching events and we’ll share with you the most interesting ones.

Adam Surak (https://twitter.com/AdamSurak) is DevOps engineer @algolia (https://www.algolia.com/)