[
https://issues.apache.org/jira/browse/PHOENIX-4013?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Samarth Jain updated PHOENIX-4013:
----------------------------------
Summary: Use CoprocessorHConnectionFactory in our coprocessors for getting
HTable till HBASE-18359 is fixed (was: Distributed UPSERT SELECT should use
CoprocessorHConnection in HTable)
> Use CoprocessorHConnectionFactory in our coprocessors for getting HTable till
> HBASE-18359 is fixed
> --------------------------------------------------------------------------------------------------
>
> Key: PHOENIX-4013
> URL: https://issues.apache.org/jira/browse/PHOENIX-4013
> Project: Phoenix
> Issue Type: Bug
> Reporter: Samarth Jain
>
> We should be using CoprocessorHConnection so that we don't go through the RPC
> stack when the destination table is on the same region server as the source
> table. One possibility is to use the CoprocessorHConnectionTableFactory added
> in PHOENIX-3994. We would need to make sure that the factory is created in
> start() and shutdown in stop() method of the UngroupedAggregateRegionObserver
> coprocessor.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
