Speaker - Martina Iglesias Fernández

15 July 2016 – Posted by Gareth Evans

Automatic discovery of service metadata for systems at scale

Documentation is a painful yet necessary artifact of any software system. In a large microservices environment like Spotify’s, where the number of services and teams grows quickly, automation and standardization of documentation is crucial.

However, maintaining any kind of quality documentation at this scale becomes a burden easily: it is a boring task, and an artifact that usually resides in a different place other than your code.

Discover how we solved this issue by including a metadata module in our main microservices framework, Apollo. Allowing automatic discovery of existing endpoints, service configuration, outgoing and incoming calls, and deployment information at runtime. All this information is visualized in our centralized systems management tool. Making our systems’ documentation easy to maintain and ensuring we always display up-to-date service metadata.

We want you to know about our open source tool, Apollo and how we solved this common, painful issue with it.


Martina is a Lead Backend Engineer at Spotify in the team responsible for album, artist and chart pages.

Before she worked as a researcher (Human Brain Project), which lead to a scientific publication. In her free time she enjoys music and dancing tango.

Talk Slides

Hypr Logo OAuth.io Logo