Health check

[shishkin]

As far as I understand, stub server only responds to requests from the pacts. I think it would make sense to also add a diagnostics route like /health or maybe better something like /_pact/health that could be used by infrastructure to know when the service is ready and if it is alive. I'm using process-compose to orchestrate dev environment and process compose can only do HTTP and command execution checks.

[rholshausen]

You could add a Pact with a single interaction that has the health route that just returns a 200 OK

[shishkin]

That would require the provider to implement that same route for verification to pass. That may not be desirable or even possible.

[rholshausen]

Why? The stub server doesn't do verification, it just there to be a pretend server.

[shishkin]

Oh, I see. I think you mean I can create a separate pact, that I don't have to verify but just feed it to the stub server. That would work.

[YOU54F]

Good little workaround Ron!

I think it is a good feature to allow it to be started with admin endpoints, that are able to be accessed with a control header, similar to how users interacted with the pact mock service (ruby version).

the rust version has a master server you interact with, and separate mock servers spun up.

I could envisage it to be useful to query the stub server to find about which pacts / interactions are loaded, metrics on matches / misses etc