Code quality: rescue Exception catches all errors including SignalException—prevents clean shutdown

[tkadauke]

Description

app/models/health_check.rb line 156 catches Exception (the base class of all Ruby exceptions):

rescue Exception => e
  retry_times -= 1
  retry unless retry_times == 0

Exception is the superclass of SignalException, SystemExit, Interrupt, NoMemoryError, and ScriptError. Catching these prevents:

• Ctrl+C (Interrupt) from stopping the server process gracefully.
• SIGTERM (SystemExit) from stopping the process—critical for deployment and container shutdown.
• Out-of-memory conditions from propagating to the process manager.

Additionally, the retry logic retries SignalException 3 times before finally re-raising... actually it never re-raises: after 3 retries it falls through to attrs[:status] = 'failure', swallowing the signal entirely.

Suggested approach

Catch only application-level errors:

rescue StandardError => e
  retry_times -= 1
  retry unless retry_times == 0
  # ...
end

Also consider rescuing specific known exceptions (Capybara::Error, Net::ReadTimeout, etc.) and handling them differently (retry on timeout, fail fast on element-not-found).

Effort: small