Monitoring tools for your microservices are the key factor to success in a real-life migration.
We worked as part of a bigger team to insure a slow but steady migration towards a micro-service architecture of an existing monolith.
When dealing with old monoliths, and want to extract in dedicated self contained micro-services you need to be sure that during the migration phase, all systems are up and running and no transactions are lost. Build infrastructure before hand, have feature flags implemented and event driven queues in place. In case something goes bad, be sure you can handle that by replaying all events and recompute the states.
All payments had to be secured. To be sure our implementations were right we made staging infrastructure deployment easy. Blue green architecture was a key factor of success in this case.
Those queues handlers and service migrations were great! Kudos!