[
https://issues.apache.org/jira/browse/HADOOP-2188?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12543650
]
Owen O'Malley commented on HADOOP-2188:
---------------------------------------
It would be better if the server stopped reading the incoming queue, but I
believe it is not critical for a first pass. The reason is that without the
client side timeouts, the size of the queue is bounded by the number of client
threads calling the server. So even if you have 2k servers running 4 maps each
with 5 threads each, that is still only 40k pending calls, which won't be
enough to blow out memory.
> RPC should send a ping rather than use client timeouts
> ------------------------------------------------------
>
> Key: HADOOP-2188
> URL: https://issues.apache.org/jira/browse/HADOOP-2188
> Project: Hadoop
> Issue Type: Improvement
> Components: ipc
> Reporter: Owen O'Malley
> Assignee: Owen O'Malley
> Fix For: 0.16.0
>
> Attachments: rpc-to.patch
>
>
> Current RPC (really IPC) relies on client side timeouts to find "dead"
> sockets. I propose that we have a thread that once a minute (if the
> connection has been idle) writes a "ping" message to the socket. The client
> can detect a dead socket by the resulting error on the write, so no client
> side timeout is required. Also note that the ipc server does not need to
> respond to the ping, just discard it.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
