fiber_error_supervision

🎯 Supervise and Propagate Errors Across Fibers

Implement a supervisor fiber to catch and route exceptions from child fibers, allowing you to restart or log failures. This pattern ensures your system remains responsive even if individual fibers crash.

class FiberSupervisor
  def initialize
    @workers = []
  end

  def spawn(&block)
    f = Fiber.new do
      begin
        block.call
      rescue => e
        puts "Worker failed: #{e.class} - #{e.message}"
        # Optionally restart or escalate
      end
    end
    @workers << f
    f.resume
  end

  def run
    loop { @workers.reject!(&:dead?) }
  end
end

supervisor = FiberSupervisor.new
supervisor.spawn { raise "Boom" }
supervisor.spawn { puts "All good" }
supervisor.run