Refactor tfchain Billing logic

[sameh-farouk]

There are known issues regarding billing.

• During the grace period, the NU consumption may be incorrectly accumulated tfchain#994
• certified capacity paid as DIY capacity  tfchain#991
• Refactor/remove lock balance logic in billing flow tfchain#990
• Contract is not getting billed tfchain#982
• What will happen to a rent contract on a grace period for an offline node  tfchain#979
• Misuse of balance lock tfchain#969
• Error distributing cultivation rewards tfchain#834
• Improve off-chain worker logic tfchain#932

This story aims to improve the logic of contract billing, making it less error-prone, and more auditable and readable. To achieve this goal, we need to

• Instrument billing logic to improve audibility and log all major steps
• Replacing the locking mechanism with a more reliable system
• Improve off-chain worker logic
  

[sameh-farouk]

Update:
The billing changes are now quite stable. I’ll begin working on complementary tasks such as adding new tests, updating documentation, and planning the necessary changes for the SDK.

I switched the PR to “ready for review” to indicate that no major changes are expected from this point forward.

[sameh-farouk]

Update:

• Added some new tests and a new ADR document.
• Resolved the comments on the PR.
• I made some minor changes, mostly hardening and adjusting function names for consisntency.
• I'm still awaiting feedback regarding whether a certified capacity should cost 25% more compared to a DIY capacity (It is one of the fixes implemented within the refactor.).

[sameh-farouk]

Update:

• Resolved all comments on the PR.
• Implemented additional test cases and ran an instance of a forked mainnet chain for several days to check for any inconsistent states.
• I have implemented and tested a regular storage migration (as opposed to the lazy one we previously implemented) to make it easier to handle legacy contract locks in the SDKs.
• Discovered a bug in the Polkadot-sdk v1.0.0 that we are currently based on, so we will probably push forward the upgrade to at least Polkadot-sdk v1.1.0 PR.
• Although the pull request is ready, there is a new requirement regarding reward distribution that awaits confirmation. If accepted, it would extend the current work by maybe a day or two.