[GitHub] [hadoop-ozone] hanishakoneru edited a comment on issue #742: HDDS-3217. Datanode startup is slow due to iterating container DB 2-3 times.

[GitBox <gi...@apache.org>]

hanishakoneru edited a comment on issue #742:
URL: https://github.com/apache/hadoop-ozone/pull/742#issuecomment-618121489


   > But anyway I beleive our recommendation will be to use FILE_PER_BLOCK going forward. To support older layout version this is being done from my understanding how we have done in this PR.
   
   In that case, why don't we go with FILE_PER_BLOCK approach and discard the FILE_PER_CHUNK going forward?
   
   Also LayoutVersions should not be configurable. In my understanding, they should only be used when there is an incompatible change. Specifically with this change, this is not really a user configurable parameter. This version onwards metadata will always be stored in DB irrespective of what value "ozone.scm.chunk.layout" is configured to.
   
   > In that case, why don't we go with FILE_PER_BLOCK approach and discard the FILE_PER_CHUNK going forward?
   
   Yes, we can remove it if not needed. But I don't want to touch this in this Jira. I can open a new jira and if we get consensus we can remove it.
   
   
   `Specifically with this change, this is not really a user configurable parameter. This version onwards metadata will always be stored in DB irrespective of what value "ozone.scm.chunk.layout" is configured to.`
   
   The reason for not removing the older version is for this, which i have mentioned in the code comments.
   
   ```
         // If on newer version of software if user configured mistakenly to use
         // older version, log an warning message. Not changing to newer layouts
         // for now to give an option to user if some bug present in newer layout
         // version, user has option to use still older version. We can change
         // this automatically in newer releases if no issues with newer version.
   ```
   
   This is done for being on safer side. 


----------------------------------------------------------------
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



---------------------------------------------------------------------
To unsubscribe, e-mail: ozone-issues-unsubscribe@hadoop.apache.org
For additional commands, e-mail: ozone-issues-help@hadoop.apache.org