Question

A network automation service written in Python runs in production and pushes config changes via RESTCONF. During normal operation the on-call team is overwhelmed by log volume, but when a rare failure occurs they lack the detailed variable values needed to troubleshoot. The developer wants a logging strategy where routine operational events are recorded compactly, while detailed diagnostic data (payloads, variable state) can be captured on demand without a code change. Which approach best meets these requirements?

Accepted Answer

Emit routine events at INFO and detailed diagnostic data at DEBUG, then set the logger level via an environment variable so DEBUG can be enabled at runtime without redeploying.

Answer

Log everything at INFO level and use grep on the syslog server to filter out the noisy messages when troubleshooting is not needed.

Answer

Wrap all diagnostic print statements in a commented-out block and have an engineer uncomment and redeploy the code whenever detailed logs are required.

Answer

Send all log records at ERROR level so only failures are stored, and reproduce issues manually in a lab to obtain variable values.