I am aware of the possibility to chose from up to 6 log levels for my RF tests as well as to choose which level should be shown by default in my HTML reports.
The question is, why wouldn’t I want to always collect logs at trace level when I can easily filter the level of logs that are being shown?
Thank you for your explanation (I hope the cause will be a bit deeper than “trace logs eat more drive space” )
Robot itself logs a lot on trace level, including all keyword arguments and return values. This can increase the output.xml and log.html size considerably. In worst cases a single huge argument or a return value (e.g. a binary file read from the file system) can make the output.xml so big that processing it fails.
Another problem is that because all argument values are logged, sensitive information can leak.
Thank you, Pekka.
Security risk did not come to my mind at all, while it’s quite an important aspect and should be always considered.
Have a nice day & weekend everyone!