logging
logging:
level: info
format: text
#disable_timestamp: true
logging.level
Default: info ReloadableControls the verbosity of logs. The options are trace, debug, info, warn (or warning), and error. fatal
and panic are accepted as aliases for error.
Debug and trace logging can log remotely controlled, untrusted data, which can quickly fill a disk in some scenarios. They are also CPU intensive and decrease overall performance. Only enable them while actively investigating an issue.
The level can also be changed at runtime with the log-level debug command. A
config reload sets it back to the configured value.
logging.format
Default: text ReloadableControls the logging format. The options are json or text.
logging.disable_timestamp
Default: false ReloadableDisables timestamp logging. Useful when output is redirected to logging system that already adds timestamps.
Timestamps are always RFC3339. The text format uses millisecond
precision, e.g. 2026-10-28T14:03:07.120-04:00, and the json format uses up to nanosecond precision with trailing
zeros dropped, e.g. 2026-10-28T14:03:07.12034567-04:00. logging.timestamp_format was removed in v1.11.0.
Rate limited messages
Some errors that can repeat for every packet are logged at most once per second each. The next line logged carries a
suppressed field with the number of lines skipped. These are:
Failed to write outgoing packetFailed to WriteTo in sendViaDropping outbound packet, out buffer not large enough