Fix/circuit breaker

[Umang01-hash]

Pull Request Template

Bug:

• When a user configured HealthConfig with a custom health endpoint, the circuit breaker still used the default /.well-known/alive endpoint for recovery checks. If this default endpoint didn't exist on the downstream service, the circuit breaker would never recover (stay permanently open).

Solution:

• The circuit breaker now automatically uses the same health endpoint configured in HealthConfig for its recovery checks. The user only needs to configure the health endpoint once.

Checklist:

• I have formatted my code using goimport and golangci-lint.
• All new code is covered by unit tests.
• This PR does not decrease the overall code coverage.
• I have reviewed the code comments and documentation for clarity.

Thank you for your contribution!

[aryanmehrotra]

You are currently updating the circuit breaker configuration inside the health-check, which is not a good practice.

At present, the HTTP configuration hierarchy looks like this:

Level 1: HTTP

Level 2: Individual Options

Each option at Level 2 is expected to be independent and should not depend on or modify another option. However, the current implementation violates this principle by allowing one option (health-check) to influence another (circuit breaker).

If a particular option requires a value that is also needed by another option, that value should be owned and managed by the parent HTTP configuration, not by one of the child options. This keeps the responsibilities clearly separated and maintains proper encapsulation.

With the current approach, if we introduce a new option in the future that also depends on health-check behavior, we would again need to modify the health-check configuration. This makes the design harder to extend and increases coupling.

Moving the shared configuration to the HTTP parent level will make the system more modular, easier to reason about, and more scalable for future extensions.