keith-turner commented on PR #3230:
URL: https://github.com/apache/accumulo/pull/3230#issuecomment-1538760161

   > I'm confused by the description I guess - is the goal to remove invalid 
locks or is the goal to stop processes running older versions of software and 
you are doing it by checking the lock data?
   
   I would say the overall goal should be to prevent the upgrade code from 
causing corruption of user metadata or data.  Root and metadata tablets running 
on old server processes while the upgrade process runs could cause corruption 
of their metadata.
   
   > I'd rather see this be a hard failure when there are old servers running. 
From a user perspective, I would rather know that I need to stop old servers or 
manually delete their ZK paths before proceeding with an upgrade.
   
   I agree with this, I think detecting old server processes and failing to 
upgrade is preferable to automatically killing them.  What if someone 
accidentally runs the new upgrade code on the wrong cluster and they would not 
have wanted it to kill all the server processes?


-- 
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.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]

Reply via email to