In recent years, there has been a push in the tech community to move away from using terms like "master" and "slave" due to their association with oppressive and discriminatory contexts. In the context of database architecture, you can replace "master" and "slave" terminologies with more neutral and descriptive terms such as: Primary - Secondary, or Primary - Replica.

OpenTelemetry is designed for busy production environments and is "battle-tested" by many large organizations. There is a minor impact on performances, you can read more about the benchmarks here: https://opentelemetry.io/docs/reference/specification/performance-benchmark/ . In our tests, we saw an impact of less than 1% in the overall avg query duration. For a super busy production DB, it probably makes sense to sample the queries, say 1 out of 20.