[
https://issues.apache.org/jira/browse/HADOOP-7557?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13126214#comment-13126214
]
Doug Cutting commented on HADOOP-7557:
--------------------------------------
Todd, I agree that we should switch to an RPC system that uses but a single
serialization system. This header format could go into a new
ProtobufRpcEngine, which determines headers, handshakes, message formats, etc.
If Sanjay comes up with a fully-functional implementation of ProtobufRpcEngine
before anyone else (like me) comes up with a fully-functional AvroRpcEngine
then we should switch to that. It's probably silly to work on both.
But if one were to design a client/server container to support messages
serialized with different serialization systems (as, e.g., HTTP is) then I
think it would be an abstraction violation to use a particular message
serialization system to define its headers. But you're right that we probably
don't need to design such a meta-container in the first place. Thrift already
defines various containers, as does Avro. Protobuf does not, so if protobuf is
selected for message serialization then we'll need to design a
protobuf-specific client/server container protocol, implement it in different
languages, etc. But it need not also support Writable or Avro.
> Make IPC header be extensible
> -------------------------------
>
> Key: HADOOP-7557
> URL: https://issues.apache.org/jira/browse/HADOOP-7557
> Project: Hadoop Common
> Issue Type: Improvement
> Reporter: Sanjay Radia
> Assignee: Sanjay Radia
> Attachments: HADOOP-7557.patch, IpcHeader.proto, ipcHeader1.patch,
> ipcHeader2.patch
>
>
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators:
https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
