bshashikant commented on pull request #2881: URL: https://github.com/apache/hadoop/pull/2881#issuecomment-819454476
> > I think We should hold this off, I think the code has issues, as I said. Earlier I thought, there is some catch but I don’t think. It is misbehaving only. Ideally such features should go in a branch first.... > > @ayushtkn , IMO its not a misbehaviour here. Once the snapshotTrash feature is enabled, the .Trash directory has to be present to make the feature work. As you can have pre existing snapshottable directories , one way to create the .Trash inside these was to create right after the startup is done. The other solution was to explicitly provision the .Trash using a cmd line option. The choice was made to do it automatically on restart, and fail the Namenode in case any any issues occur. > > The other solution is not fail the namenode , but log an warning and later, if any Trash operations gets performed inside the snapshottable root, will fail. > > The discussion is here: [#2682 (comment)](https://github.com/apache/hadoop/pull/2682#discussion_r570461526) Coming to think of it, if providing an external command to create the Trash directory by admins is feasible and makes sense, i think its ok to remove the NN startup logic to create Trash directories inside snapshottable root. -- 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: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
