[ 
https://issues.apache.org/jira/browse/STORM-297?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14020872#comment-14020872
 ] 

ASF GitHub Bot commented on STORM-297:
--------------------------------------

Github user clockfly commented on the pull request:

    https://github.com/apache/incubator-storm/pull/103#issuecomment-45414524
  
    @Bobby, I reproduced this. It is random. I will look into this, thanks.
    
    Seems the Netty Client is closed before worker transfer thread is closed.
    So, transfer thread got an exception.
    
    137917 [Thread-549-disruptor-worker-transfer-queue] ERROR 
backtype.storm.util - Async loop died!
    java.lang.RuntimeException: java.lang.RuntimeException: Client is being 
closed, and does not take requests any more
        at 
backtype.storm.utils.DisruptorQueue.consumeBatchToCursor(DisruptorQueue.java:114)
 ~[classes/:na]
        at 
backtype.storm.utils.DisruptorQueue.consumeBatchWhenAvailable(DisruptorQueue.java:85)
 ~[classes/:na]
        at 
backtype.storm.disruptor$consume_batch_when_available.invoke(disruptor.clj:78) 
~[classes/:na]
        at 
backtype.storm.disruptor$consume_loop_STAR_$fn__2895.invoke(disruptor.clj:89) 
~[classes/:na]
        at backtype.storm.util$async_loop$fn__442.invoke(util.clj:434) 
~[classes/:na]
        at clojure.lang.AFn.run(AFn.java:24) [clojure-1.4.0.jar:na]
        at java.lang.Thread.run(Thread.java:679) [na:1.6.0_22]
    Caused by: java.lang.RuntimeException: Client is being closed, and does not 
take requests any more
        at backtype.storm.messaging.netty.Client.send(Client.java:194) 
~[classes/:na]
        at backtype.storm.utils.TransferDrainer.send(TransferDrainer.java:54) 
~[classes/:na]
        at 
backtype.storm.daemon.worker$mk_transfer_tuples_handler$fn__5765$fn__5766.invoke(worker.clj:322)
 ~[classes/:na]
        at 
backtype.storm.daemon.worker$mk_transfer_tuples_handler$fn__5765.invoke(worker.clj:320)
 ~[classes/:na]
        at 
backtype.storm.disruptor$clojure_handler$reify__2882.onEvent(disruptor.clj:59) 
~[classes/:na]
        at 
backtype.storm.utils.DisruptorQueue.consumeBatchToCursor(DisruptorQueue.java:111)
 ~[classes/:na]
        ... 6 common frames omitted



> 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_conf.txt, 
> storm_performance_fix.patch, worker_throughput_without_storm-297.png
>
>
> 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)

Reply via email to