Interview with Paul Dix

QCon: What is a time series database?

Paul: It’s basically a database where everything is indexed by time. When most people think of a time series database, they are thinking about databases organized around regular time series data (which is samples at fixed intervals of time). But a true time series database should be able to support regular and irregular data. Irregular data is like event data, such as pageviews or user actions in an app or maybe trades in a stock market.

QCon: What made you choose Go?

Paul: I was initially interested in the language. So I spiked with it back in 2012 (over the course of a week), and I was impressed with the performance. We’ve been using it ever since.

QCon: What are some of the things you will speak about in your talk?

Paul: We started with Go in 2012. I will talk about our experiences over that time. I will talk about some of the benefits of Go, such as the community and performance. I will also talk about some of the challenges we’ve had implementing Go. For example, we’ve had a couple of issues that are still outstanding with migrating to 1.5. We actually may need to revert to 1.4, so I will speak about those things.

I will talk quite a bit about dependency management. Dependency management is one of those things over the lifetime of the project that has caused us quite of bit of bugs, so you want to get dependency management right from the start.