[
https://issues.apache.org/jira/browse/HADOOP-11552?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Siddharth Seth updated HADOOP-11552:
------------------------------------
Attachment: HADOOP-11552.1.wip.txt
Work in progress patch to get input on the approach.
Trying to make this as unobtrusive as possible so that APIs do not break.
Ofcourse, changing YARN, HDFS, etc to use something like this will require a
fair amount of attention to ensure compatibility.
Individual methods to decide whether they want to perform the operation within
the handler thread, or send the response back at a later point. Borrows
elements from HADOOP-10300.
There's several bits pending - like error handling.
Also, this ends up sending the response back on the application thread which
invokes the reply - which may not work for models like YARN where the central
dispatcher would end up blocking. Ideally, we should be able to use the handler
threads to send these responses. Alternately, with the current patch - a
threadpool for sending responses could be setup.
Additional considerations - number of outstanding requests, and whether they
need to be limited at the RPC layer.
> Allow handoff on the server side for RPC requests
> -------------------------------------------------
>
> Key: HADOOP-11552
> URL: https://issues.apache.org/jira/browse/HADOOP-11552
> Project: Hadoop Common
> Issue Type: Improvement
> Components: ipc
> Reporter: Siddharth Seth
> Attachments: HADOOP-11552.1.wip.txt
>
>
> An RPC server handler thread is tied up for each incoming RPC request. This
> isn't ideal, since this essentially implies that RPC operations should be
> short lived, and most operations which could take time end up falling back to
> a polling mechanism.
> Some use cases where this is useful.
> - YARN submitApplication - which currently submits, followed by a poll to
> check if the application is accepted while the submit operation is written
> out to storage. This can be collapsed into a single call.
> - YARN allocate - requests and allocations use the same protocol. New
> allocations are received via polling.
> The allocate protocol could be split into a request/heartbeat along with a
> 'awaitResponse'. The request/heartbeat is sent only when there's a request or
> on a much longer heartbeat interval. awaitResponse is always left active with
> the RM - and returns the moment something is available.
> MapReduce/Tez task to AM communication is another example of this pattern.
> The same pattern of splitting calls can be used for other protocols as well.
> This should serve to improve latency, as well as reduce network traffic since
> the keep-alive heartbeat can be sent less frequently.
> I believe there's some cases in HDFS as well, where the DN gets told to
> perform some operations when they heartbeat into the NN.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
