[
https://issues.apache.org/jira/browse/HBASE-7898?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13583470#comment-13583470
]
stack commented on HBASE-7898:
------------------------------
Working on this issue, did some study of rpc. Here are findings, why some
decisions, and list of TODOS:
https://docs.google.com/document/d/1WCKwgaLDqBw2vpux0jPsAu2WPTRISob7HGCO8YhfDTA/edit#
Here is a specification for wire-format:
https://docs.google.com/document/d/1-1RJMLXzYldmHgKP7M7ynK6euRpucD03fZ603DlZfGI/edit#
Needs polish but basically there. This can be used to get high-level picture
when trying to situate the patches that will be coming in as subtasks of this
jira.
> Serializing Cells over RPC
> --------------------------
>
> Key: HBASE-7898
> URL: https://issues.apache.org/jira/browse/HBASE-7898
> Project: HBase
> Issue Type: Task
> Components: IPC/RPC
> Reporter: stack
> Assignee: stack
> Priority: Blocker
> Fix For: 0.96.0
>
>
> This issue replaces the juicy HBASE-7233 which started undoing KeyValue as a
> Writable and ended up as a general discussion of KeyValue serialization.
> Along the way it served as a scratch sheet for how to do lots of KeyValues
> across RPC. This issue is the umbrella issue for making it so we can pass
> Cells over our protobuf'd RPC rather than KeyValues; i.e. an Interface which
> could be differently implemented rather than our current non-evolvable
> KeyValue serialization.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
