[EPIC] Unified Log Output
It would make debugging a lot easier if all the logs had the same structure and format. Here are some points I think we could improve:
-
unified format on the log-line prefix level: same date/time format, same position of the service/component identifier allowing monitoring software to watch specific log events. For example 2024-03-21 12:42:23,387:engine:ERROR - %MESSAGE%
-
unified log levels. Use debug, info, warning, error, critical
, instead of5,4,3,2,1
(as inengine-core
). Default should beinfo
. Check other services for having a different level scheme. -
printing with colors and handling colors for files: Use green for important, successful information, red for errors. -
unified user rights for log files, for example 664
: This can be very different from service to service e.g. Nginx. -
unified startup message. Print version
, welcome banner or something like a horizontal line indicating that the startup is successfully finished. This is helpful when several service are observed usingdocker compose logs
.
Maybe we could agree on some standards and implement them in all components.
Edited by David Trattnig