capistrant commented on issue #6873: Kill Segment Grace Period
URL: 
https://github.com/apache/incubator-druid/issues/6873#issuecomment-478030751
 
 
   @jihoonson 
   
   `druid.coordinator.kill.durationToRetain` is useful for us in some cases. We 
currently use it so protect the last N days from being killed (usually 7 or 14 
days for us).
   
   The use case I am referring to is more of a `soft kill` for a set period fo 
time followed by an actual kill if no action was taken to re-enable the 
segments. This opposed to a permanent retain rule like the `durationToRetain` 
config does. Our initial idea was to only have the kill task runner run once 
per day, but that only gives us a good grace period for segments that wen't to 
unused shortly after the previous execution of the kill task process. Our main 
problem has been disabling datasources that were still needed and having them 
be killed before we were alerted to the problem.
   
   Now that I am thinking about this more, it may just be an administrative 
issue that my org needs to sort out on our own instead of adding this guard. We 
run a multi-tenant cluster in production with a large number of end users so 
the lines on what data is useful and what is stale can get blurred for us. Long 
story short, maybe this proposal isn't the best solution to the problem we face.

----------------------------------------------------------------
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]


With regards,
Apache Git Services

---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to