Not a member of Pastebin yet?
Sign Up,
it unlocks many cool features!
- # Healthcheck
- ## Classification of health check errors
- ### Transient
- Temporary errors due to some networking issues or load could be handled by the module itself through retries, processing request queue, emptying connection pool
- and etc.
- As errors could lead to inconsistent state or unavailability, requests to the service should be avoided until the problem is fixed by module
- Examples:
- A database connection is broken, a connection pool is full, message broker connection is broken for a long time
- ### Manageable
- Some errors that are not related to infrastructure problems but happen to be a result of some implementation or configuration errors in third-party service. Still could be fixed by module through retries.
- As this kind of errors could impact all instances of the service it is very dangerous to avoid requests or stop service as it could lead to overall unavailability
- Examples:
- Configuration update parsing, external service response parsing error
- ### Fatal
- Some errors could not be fixed by module, lead to service unavailability or to unaffordable inconsistent state.
- Examples:
- Initial state load error, connection to a database could not be established due to a configuration error, code unexpected errors
Add Comment
Please, Sign In to add comment