Prometheus metrics at commands level

[vseoane96]

This library provides awesome features and functionalities and it is probably the most powerful Go client for Redis. However, in terms of observability, I can see there is only a prometheus collector for pool metrics. I am missing metrics at the commands level, which are very useful when troubleshooting and analysing the operations taking place in the Redis:

• Histogram of duration of the commands
• Counter of errors

Even if there is some other project adding the metrics through a hook, I'd like to discuss if this is something planned to be added to the project or if it has been discarded for any reason.

[vmihailenco]

There are OpenTelemetry metrics that can be exported to Prometheus: https://redis.uptrace.dev/guide/go-redis-monitoring.html

[a5r0n]

Can we use this without otel collector? Just the Prometheus exported on the http port?