feat: isolate components

[AquiGorka]

These changes introduce a handful of new features. From the top:

Local setup

To run a local Lilypad network a local blockchain, a solver service, a resource provider service, a bacalhau node and a job creator service are required.

First, start the blockchain from scratch: ./stack chain-docker-build and ./stack chain-docker-run ./path-to-local-folder-for-data, then ./stack chain-fund-admin and finally ./stack chain-boot (fund accounts, and compile and deploy contracts). The docker container includes a handy ./reset script to go back to a clean slate (it funds the admin account so you can skip that step if reset is used).

Second, start a solver service: ./stack solver (or ./stack solver-docker-build and ./stack solver-docker-run). The output here will let us know that the solver's server address has been set in the blockchain (for local dev this is http://localhost:8080). Which is required for the mediator, resource-provider and job-creator services as these all connect to the solver via a web socket.

Third, start a job creator service: ./stack job-creator (or ./stack job-creator-docker-build and ./stack job-creator-docker-run).

Fourth, start a resource provider service: ./stack resource-provider.

Fifth, start a bacalhau node: ./stack bacalhau-node

• These last two are under development, as I'm still working on being able to execute the bacalhau node inside a container (it needs to run in a container that has a docker daemon - containers inside containers); when I can successfully run a job with the node running in a container, then I plan to pack together the resource-provider and bacalhau into 1 image, as these 2 are required to setup a node to join the network.

Sixth, trigger an onchain job: ./stack run-cowsay-onchain and see how the whole flow works out.

• This last item still needs the doppler cmd to be put in place.

Remaining

• mediator this service is not required to run a job. I've yet to implement the Dockerfile and cmds to run this task
• integration-tests I've yet to set the doppler cmd in place to run this.
• all work related to the faucet task
• add cloudflare tunnels to expose the services that should be public
• github actions to deploy the whole infra
  • fix the tests
• diagram the flow task
• Update the CONTRIBUTING instructions (and potentially remove the TESTNETDEPLOY file) task

Notes

• No need to rebuild the images if there are no changes to the code. It is faster to run the non containerized components if there are changes to be tested but for the components that don't have changes (since the images use the built code).
• I highly suggest reading through the commits in order to review the code.

[10d9e]

LGTM