[
https://issues.apache.org/jira/browse/STORM-297?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14006149#comment-14006149
]
ASF GitHub Bot commented on STORM-297:
--------------------------------------
Github user nathanmarz commented on the pull request:
https://github.com/apache/incubator-storm/pull/103#issuecomment-43917540
I would like to see a test added that tests that messages are received
between tasks in the same order they are sent. The test is likely to be
probabilistic, in that if the code is wrong it won't always fail, but that's
ok. This is a really important property to maintain that this patch, or future
modifications on this code, could affect. So it needs better testing. The test
should have many workers with many send/receive threads.
> Storm Performance cannot be scaled up by adding more CPU cores
> --------------------------------------------------------------
>
> Key: STORM-297
> URL: https://issues.apache.org/jira/browse/STORM-297
> Project: Apache Storm (Incubating)
> Issue Type: Bug
> Reporter: Sean Zhong
> Labels: Performance, netty
> Fix For: 0.9.2-incubating
>
> Attachments: Storm_performance_fix.pdf,
> storm_Netty_receiver_diagram.png, storm_performance_fix.patch
>
>
> We cannot scale up the performance by adding more CPU cores and increasing
> parallelism.
> For a 2 layer topology Spout ---shuffle grouping--> bolt, when message size
> is small (around 100 bytes), we can find in the below picture that neither
> the CPU nor the network is saturated. When message size is 100 bytes, only
> 40% of CPU is used, only 18% of network is used, although we have a high
> parallelism (overall we have 144 executors)
--
This message was sent by Atlassian JIRA
(v6.2#6252)
