We should add more useful metrics:
- Lag behind master
- Number of TX
- Position in source log (currently part of the "event" string, but it should be a separate field; atm. we call toString() on the SourceInfo struct, which kinda works, but exposes some redundant information. Instead, there should be a dedicated method OffsetContext#asLogString() for this purpose)
- Actual size of internal event queue (if it's constantly near to its max capacity, it indicates some issue)
These should also be exposed in the Grafana dashboard we provide in the examples project. Related to this, the dashboard should also be able to expose all the additional metrics specific to the MySQL connector.