[
https://issues.apache.org/jira/browse/FLINK-930?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14033500#comment-14033500
]
ASF GitHub Bot commented on FLINK-930:
--------------------------------------
Github user rmetzger commented on a diff in the pull request:
https://github.com/apache/incubator-flink/pull/19#discussion_r13845310
--- Diff:
stratosphere-runtime/src/main/java/eu/stratosphere/runtime/io/network/netty/NettyConnectionManager.java
---
@@ -108,9 +130,9 @@ public void initChannel(SocketChannel channel) throws
Exception {
.option(ChannelOption.SO_KEEPALIVE, true);
try {
- this.in.bind().sync();
+ in.bind().sync();
} catch (InterruptedException e) {
- throw new RuntimeException("Could not bind server
socket for incoming connections.");
+ throw new IOException("Interrupted while trying to bind
server socket.");
--- End diff --
I would also forward the exception here
> Netty Initialization is sometimes very slow
> -------------------------------------------
>
> Key: FLINK-930
> URL: https://issues.apache.org/jira/browse/FLINK-930
> Project: Flink
> Issue Type: Bug
> Reporter: Stephan Ewen
> Assignee: Ufuk Celebi
> Labels: performance
>
> I often see the initialization of Netty taking 3 seconds and then failing
> with the message
> "WARN io.netty.util.internal.ThreadLocalRandom - Failed
> to generate a seed from SecureRandom within 3 seconds. Not enough entrophy?"
> This adds to every single run a 3 seconds delay, which is often more than the
> test time and causes our builds to run out of time.
--
This message was sent by Atlassian JIRA
(v6.2#6252)
