2 posts tagged with "guides"

While support for logging in the OpenTelemetry Python project is listed as 'experimental,' it's completely possible to send logs from your Python application. The Opentelemetry Collector has support for numerous existing logging systems, effectively exporting log data from wherever you were sending logs currently; you can also use the filelog receiver to tail and send logs from files. The only 'experimental' portion of the Python SDK is sending logs directly from code-level instrumentation.

This guide is for anyone who is getting started monitoring their application with OpenTelemetry, and is generating unstructured logs. As is well understood at this point, structured logs are ideal for post-hoc incident analysis and broad-range querying of your data. However, it’s not always feasible to implement highly structured logging at the code level.

With SigNoz, you get some parsing automatically to identify details like timestamp, container ID, container name, and an optional body. But it’s possible to go much deeper with a relatively simple configuration. It’s also a good idea to check for attributes that may contain Personal Identifying Information (PII) and remove them with filters. Since the SigNoz collector is a fork of the OpenTelemetry collector, this tutorial will also work for configuring a baseline OpenTelemetry collector.