You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@phoenix.apache.org by "James Taylor (JIRA)" <ji...@apache.org> on 2017/10/18 22:48:00 UTC

[jira] [Created] (PHOENIX-4301) Use operation.setPriority() instead of having our own PhoenixRpcSchedulerFactory

James Taylor created PHOENIX-4301:
-------------------------------------

             Summary: Use operation.setPriority() instead of having our own PhoenixRpcSchedulerFactory
                 Key: PHOENIX-4301
                 URL: https://issues.apache.org/jira/browse/PHOENIX-4301
             Project: Phoenix
          Issue Type: Bug
            Reporter: James Taylor


Now that it's possible to set the priority of an operation, we can get rid of the usage of PhoenixRpcSchedulerFactory and just set the priority ourselves since the RS->RS calls are in a known set of limited locations:
- ParallelWriterIndexCommitter {{table.batch()}} call
- TrackingParallelWriterIndexCommitter {{table.batch()}} call
- Invocation of MetaDataEndPointImpl calls such as the one that updates the index status in PhoenixIndexFailurePolicy
- Get issued while running stats to get guidepost width in DefaultStatisticsCollector

The last two aren't as important as the first two. If feasible, we can make all interactions with SYSTEM.CATALOG and MetaDataEndPointImpl higher priority.



--
This message was sent by Atlassian JIRA
(v6.4.14#64029)