How do you implement fault tolerance using try-except blocks in code?

Fault tolerance in Python is often achieved by wrapping risky operations in try-except blocks. Inside the try block, I perform the operation that may raise an exception (e.g., network call, file I/O). In the except clause, I catch specific exception types to avoid swallowing unrelated errors. I log the exception details using the logging module, which helps with debugging and monitoring. After logging, I implement a retry strategy: either a simple loop with a delay or a library like tenacity that provides exponential backoff and jitter. If the operation is idempotent, I can safely retry without side effects. For non‑idempotent actions, I use compensating actions or idempotency keys. Additionally, I may raise a custom exception after a maximum number of retries to signal failure to higher layers. This pattern ensures that transient failures are handled gracefully while still surfacing critical errors for further investigation.