Observability of Microfrontends

We have almost 1,000 microfrontends in Avito, and we need to monitor them somehow. We've designed an approach to microfrontend observability that eliminates the need to build everything from scratch for each service, and allows us to expand the system as needed.

The main ideas of speech:

• what are microfrontends and why do they require special monitoring;
• what our observability system consists of;
• what tools do we use for logs, metrics, and error logging;
• what do we log, what metrics do we send, and why;
• how do we extend best practices to multiple microfrontends.