[GitHub] [accumulo] dlmarion edited a comment on pull request #2422: Implementation of Scan Servers.

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

dlmarion edited a comment on pull request #2422:
URL: https://github.com/apache/accumulo/pull/2422#issuecomment-1021503545


   @keith-turner and I had a discussion about changing the current design. The key points from the discussion are:
   
   1. Modifying the load balancing logic from the current implementation (which uses ZK to know which scan servers are available) to an implemenation that selects  candidate scan servers to communicate with for scanning an extent. This new implementation would not require ZK and would use a hashing algorithm so that all clients select the same set of scan servers for an extent in an attempt to make use of the index and data block caches.
   
   2. We may want to have configuration settings for the scan server for things that are inherited from the tablet server. For example, the data block cache size. Does the scan server use the current configuration property name, or should it have its own.
   
   3. ~~Modifying the scan server to perform more than one concurrent scan.~~ Added in [0d553de]
   
   4. ~~Modify the client configuration code such that the name is not tied to the implementation.~~ (This has been completed with the addition of `ScannerBase.setConsistencyLevel(ConsistencyLevel)` 


-- 
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: notifications-unsubscribe@accumulo.apache.org

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