You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@systemds.apache.org by GitBox <gi...@apache.org> on 2020/07/22 07:29:43 UTC

[GitHub] [systemds] Baunsgaard commented on pull request #980: [SYSTEMDS-135] log4j.properties simplified

Baunsgaard commented on pull request #980:
URL: https://github.com/apache/systemds/pull/980#issuecomment-662292217


   > What exactly was changed by this? I get that the file is now much smaller, but I don't get why we had all those settings. Are they safe to delete?
   
   They should all be safe to delete, all of them are customization of where to log information to. I much prefer the simplicity and control of this simplification. The downside (in my opinion not so much) it is not leveraging some of the "fancy features of log4j", for instance you can split up different logging to different outputs such as saying Hadoop should log to `x.log` while spark to `y.log`.
   
   The one difference that could be debated is the security logging (from hadoop only nothing with out system) that is customized in the old settings, to log to the Hadoop security log.
   
   > I am also in favor of deleting of the template files.
   
   Historic reason for the template files seems to be that a user downloading the system would copy these into real files, that they could then customize for their own settings while being git ignored. Also the release scripts seems to refer to these template files, to copy them into a release as the official settings. 
   
   I think for simplicity in reproducing settings for each user all should share the same settings file, and we therefore remove the template files. 


----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org