[
https://issues.apache.org/jira/browse/ARTEMIS-994?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15906726#comment-15906726
]
Michael Andre Pearce (IG) commented on ARTEMIS-994:
---------------------------------------------------
Ive found my issue, it seems in the connector a custom bytebuf allocator exists
in artemis causing the bad address issue when using epoll.
bootstrap.childOption(ChannelOption.ALLOCATOR,
PartialPooledByteBufAllocator.INSTANCE);
When i make this not be set if epoll exists aka uses the default then it
seemingly works fine.
Before i look to use this as a solution i hope someone with some more history
can explain what or why this custom variant exists?
btw my fork/branch is updated with this now in this working state, toggling to
use the default allocator when epoll is loaded for now.
> Support Netty Native Epoll on Linux
> -----------------------------------
>
> Key: ARTEMIS-994
> URL: https://issues.apache.org/jira/browse/ARTEMIS-994
> Project: ActiveMQ Artemis
> Issue Type: New Feature
> Components: Broker
> Affects Versions: 1.5.3, 2.0.0
> Reporter: Michael Andre Pearce (IG)
>
> Netty has support for native epoll, this is available only on linux systems.
> This is more performant.
> http://netty.io/wiki/native-transports.html
> This is inline with supporting asyncio with libaio using such native feature
> to bring perfomance benefits to artemis.
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
