[ https://issues.apache.org/jira/browse/HBASE-17273?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15736310#comment-15736310 ]
stack commented on HBASE-17273: ------------------------------- [~vrushalic] FYI, master branch first is how folks usually do this. Thanks V. > Create an hbase-coprocessor module as repository for generally useful > coprocessors > ---------------------------------------------------------------------------------- > > Key: HBASE-17273 > URL: https://issues.apache.org/jira/browse/HBASE-17273 > Project: HBase > Issue Type: Task > Components: Coprocessors > Reporter: stack > Attachments: HBASE-17273.master.001.patch, > HBASE-17273.master.002.patch > > > Idea here is a module where we can carry coprocessors that are of general > utility. In particular, I am thinking of the coprocessor used by yarn > timeline server v2 which does aggregating and then replacing increments with > a sum cell. This seems generally useful. Other candidates might include our > cross-row transactions coprocessor is another that could go here pulled out > of hbase-server and so on. > Currently only coprocessors bundled in hbase are examples and endpoints. -- This message was sent by Atlassian JIRA (v6.3.4#6332)