Collect more metrics for monitoring and alerting

[linki]

The prometheus endpoint should expose more metrics that allow to detect whether stackset controller is running well.

Some possibilities:

• number of errors / error rate
• number of stacksets / stacks under control