[
https://issues.apache.org/jira/browse/FLUME-2243?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13843940#comment-13843940
]
Ashish Paliwal commented on FLUME-2243:
---------------------------------------
[~hshreedharan] This is not committed. FLUME-2238 was meant to control the
number of Threads. This patch is only meant to use the Thread Naming Factory,
which would aid in debugging in Thread dumps.
Verified in both commit message and trunk source, the changes are not present.
Let me know if it makes sense.
> AvroSource to use TransceiverThreadFactory for Thread naming while
> initializing NioServerSocketChannelFactory
> -------------------------------------------------------------------------------------------------------------
>
> Key: FLUME-2243
> URL: https://issues.apache.org/jira/browse/FLUME-2243
> Project: Flume
> Issue Type: Improvement
> Components: Sinks+Sources
> Affects Versions: v1.4.0
> Reporter: Ashish Paliwal
> Assignee: Ashish Paliwal
> Priority: Minor
> Attachments: FLUME-2243-0.patch
>
>
> Currently in Avro source, we use
> socketChannelFactory = new NioServerSocketChannelFactory
> (Executors .newCachedThreadPool(), Executors.newCachedThreadPool());
> This would create generic Thread names like pool-1-thread-1. It would be good
> to use a ThreadFactory, like we use in NettyAvroRpcClient for better
> identification of Threads in log file
> ExecutorService bossExecutor = Executors.newCachedThreadPool(new
> TransceiverThreadFactory(
> "Avro " + NettyTransceiver.class.getSimpleName() + " Boss"));
> ExecutorService workerExecutor = Executors.newCachedThreadPool(new
> TransceiverThreadFactory(
> "Avro " + NettyTransceiver.class.getSimpleName() + " I/O Worker"));
--
This message was sent by Atlassian JIRA
(v6.1.4#6159)
