[jira] Commented: (DIRMINA-762) WARN org.apache.mina.core.service.IoProcessor - Create a new selector. Selected is 0, delta = 0
[ https://issues.apache.org/jira/browse/DIRMINA-762?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12829095#action_12829095 ] Emmanuel Lecharny commented on DIRMINA-762: --- It's interesting. I doubt that you obtain 200k msg/s when doing ( connect/sendMsg/receivedResp/disconnect ) though. In a scenario where you establish a connection, send N mesg, then disconect, yes, I agree, you will have a way higher number of messages exchanged. Here, I think this is the big difference. > WARN org.apache.mina.core.service.IoProcessor - Create a new selector. > Selected is 0, delta = 0 > > > Key: DIRMINA-762 > URL: https://issues.apache.org/jira/browse/DIRMINA-762 > Project: MINA > Issue Type: Bug > Environment: Linux (2.6.26-2-amd64), java version 1.6.0_12 and also > 1.6.0_18. >Reporter: Omry Yadan >Priority: Critical > Fix For: 2.0.0-RC2 > > Attachments: BufferCodec.java, NettyTestServer.java, > RateCounter.java, Screen shot 2010-02-02 at 7.48.39 PM.png, Screen shot > 2010-02-02 at 7.48.46 PM.png, Screen shot 2010-02-02 at 7.48.59 PM.png, > Screen shot 2010-02-02 at 7.49.13 PM.png, Screen shot 2010-02-02 at 7.49.18 > PM.png, Server.java, StressClient.java > > > Mina server gets into a bad state where it constantly prints : > WARN org.apache.mina.core.service.IoProcessor - Create a new selector. > Selected is 0, delta = 0 > when this happens, server throughput drops significantly. > to reproduce run the attached server and client for a short while (30 seconds > on my box). -- This message is automatically generated by JIRA. - You can reply to this email to add a comment to the issue online.
[jira] Commented: (DIRMINA-762) WARN org.apache.mina.core.service.IoProcessor - Create a new selector. Selected is 0, delta = 0
[ https://issues.apache.org/jira/browse/DIRMINA-762?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12829113#action_12829113 ] Omry Yadan commented on DIRMINA-762: I compared apples to apples. the scenario with 64 persistent connections firing messages without any delay was run against both servers. the problem with open-send-receive-close test is that it's very easy to hit that pesky "java.net.NoRouteToHostException: Cannot assign requested address" I tweaked my linux tcp stack a bit and to better handle the open-send-receive-close use case (reuse time_wait and crap like that). the test was 3 threads a delay of 5 ms between messages. both servers had similar throughput (~560/sec for netty, ~540/sec for mina). this is nearly optimal, the maximum number of messages for this setup would be (1000/5)* 3=600. this indicates that I could get higher had I worked around the local ports issue so in any case it's not a very good te > WARN org.apache.mina.core.service.IoProcessor - Create a new selector. > Selected is 0, delta = 0 > > > Key: DIRMINA-762 > URL: https://issues.apache.org/jira/browse/DIRMINA-762 > Project: MINA > Issue Type: Bug > Environment: Linux (2.6.26-2-amd64), java version 1.6.0_12 and also > 1.6.0_18. >Reporter: Omry Yadan >Priority: Critical > Fix For: 2.0.0-RC2 > > Attachments: BufferCodec.java, NettyTestServer.java, > RateCounter.java, Screen shot 2010-02-02 at 7.48.39 PM.png, Screen shot > 2010-02-02 at 7.48.46 PM.png, Screen shot 2010-02-02 at 7.48.59 PM.png, > Screen shot 2010-02-02 at 7.49.13 PM.png, Screen shot 2010-02-02 at 7.49.18 > PM.png, Server.java, StressClient.java > > > Mina server gets into a bad state where it constantly prints : > WARN org.apache.mina.core.service.IoProcessor - Create a new selector. > Selected is 0, delta = 0 > when this happens, server throughput drops significantly. > to reproduce run the attached server and client for a short while (30 seconds > on my box). -- This message is automatically generated by JIRA. - You can reply to this email to add a comment to the issue online.
[jira] Commented: (DIRMINA-762) WARN org.apache.mina.core.service.IoProcessor - Create a new selector. Selected is 0, delta = 0
[ https://issues.apache.org/jira/browse/DIRMINA-762?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12829120#action_12829120 ] Emmanuel Lecharny commented on DIRMINA-762: --- Yeah, I did some tuning too n my laptop (setting the delay to 1 sec instead of a 15 sec defaut value) in order to be able to launch more clients. Do you have the modified client and server tests without all those connections and disconnection ? I have many servers available (on linux) right now, and I can run a test (5 injectors, one server) > WARN org.apache.mina.core.service.IoProcessor - Create a new selector. > Selected is 0, delta = 0 > > > Key: DIRMINA-762 > URL: https://issues.apache.org/jira/browse/DIRMINA-762 > Project: MINA > Issue Type: Bug > Environment: Linux (2.6.26-2-amd64), java version 1.6.0_12 and also > 1.6.0_18. >Reporter: Omry Yadan >Priority: Critical > Fix For: 2.0.0-RC2 > > Attachments: BufferCodec.java, NettyTestServer.java, > RateCounter.java, Screen shot 2010-02-02 at 7.48.39 PM.png, Screen shot > 2010-02-02 at 7.48.46 PM.png, Screen shot 2010-02-02 at 7.48.59 PM.png, > Screen shot 2010-02-02 at 7.49.13 PM.png, Screen shot 2010-02-02 at 7.49.18 > PM.png, Server.java, StressClient.java > > > Mina server gets into a bad state where it constantly prints : > WARN org.apache.mina.core.service.IoProcessor - Create a new selector. > Selected is 0, delta = 0 > when this happens, server throughput drops significantly. > to reproduce run the attached server and client for a short while (30 seconds > on my box). -- This message is automatically generated by JIRA. - You can reply to this email to add a comment to the issue online.
[jira] Commented: (DIRMINA-762) WARN org.apache.mina.core.service.IoProcessor - Create a new selector. Selected is 0, delta = 0
[ https://issues.apache.org/jira/browse/DIRMINA-762?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12831485#action_12831485 ] Emmanuel Lecharny commented on DIRMINA-762: --- Ok, I found some time to run the new client. I get some interesting results : 3092.25 messages/sec (total messages received 15235) 3231.0 messages/sec (total messages received 31367) ... 3232.25 messages/sec (total messages received 79864) Warning : Short connections thread 2 have no activity for 1322 ms Warning : Long connection thread 0 have no activity for 1271 ms Warning : Long connection thread 1 have no activity for 1262 ms Warning : Long connection thread 3 have no activity for 1271 ms Warning : Long connection thread 6 have no activity for 1264 ms Warning : Long connection thread 10 have no activity for 1269 ms Warning : Long connection thread 17 have no activity for 1262 ms Warning : Long connection thread 18 have no activity for 1261 ms Warning : Long connection thread 21 have no activity for 1271 ms Warning : Long connection thread 24 have no activity for 1267 ms Warning : Long connection thread 27 have no activity for 1271 ms Warning : Long connection thread 28 have no activity for 1261 ms Warning : Long connection thread 31 have no activity for 1262 ms 2262.5 messages/sec (total messages received 91947) 1733.0 messages/sec (total messages received 100753) Warning : Short connections thread 1 have no activity for 1479 ms Warning : Long connection thread 4 have no activity for 1421 ms Warning : Long connection thread 7 have no activity for 1430 ms Warning : Long connection thread 11 have no activity for 1422 ms Warning : Long connection thread 13 have no activity for 1424 ms Warning : Long connection thread 14 have no activity for 1430 ms Warning : Long connection thread 16 have no activity for 1422 ms Warning : Long connection thread 19 have no activity for 1424 ms Warning : Long connection thread 22 have no activity for 1423 ms Warning : Long connection thread 26 have no activity for 1430 ms Warning : Long connection thread 29 have no activity for 1421 ms 1041.25 messages/sec (total messages received 105948) ... 1035.5 messages/sec (total messages received 163086) Warning : Short connections thread 4 have no activity for 1844 ms Warning : Long connection thread 2 have no activity for 1793 ms Warning : Long connection thread 5 have no activity for 1786 ms Warning : Long connection thread 8 have no activity for 1785 ms Warning : Long connection thread 9 have no activity for 1785 ms Warning : Long connection thread 12 have no activity for 1793 ms Warning : Long connection thread 15 have no activity for 1783 ms Warning : Long connection thread 20 have no activity for 1793 ms Warning : Long connection thread 23 have no activity for 1784 ms Warning : Long connection thread 25 have no activity for 1786 ms Warning : Long connection thread 30 have no activity for 1785 ms 39.5 messages/sec (total messages received 163483) ... There are 32 long connection created, and one third (roughly) is being 'killed', or stall, then a few moment later, another third stall, and at the end, the last third stall. At this point, we don't receive any message on the long connections. Seems like we have a problem writing data to the client at some point. The selector is not swapped, btw. > WARN org.apache.mina.core.service.IoProcessor - Create a new selector. > Selected is 0, delta = 0 > > > Key: DIRMINA-762 > URL: https://issues.apache.org/jira/browse/DIRMINA-762 > Project: MINA > Issue Type: Bug > Environment: Linux (2.6.26-2-amd64), java version 1.6.0_12 and also > 1.6.0_18. >Reporter: Omry Yadan >Priority: Critical > Fix For: 2.0.0-RC2 > > Attachments: BufferCodec.java, NettyTestServer.java, > RateCounter.java, Screen shot 2010-02-02 at 7.48.39 PM.png, Screen shot > 2010-02-02 at 7.48.46 PM.png, Screen shot 2010-02-02 at 7.48.59 PM.png, > Screen shot 2010-02-02 at 7.49.13 PM.png, Screen shot 2010-02-02 at 7.49.18 > PM.png, Server.java, StressClient.java > > > Mina server gets into a bad state where it constantly prints : > WARN org.apache.mina.core.service.IoProcessor - Create a new selector. > Selected is 0, delta = 0 > when this happens, server throughput drops significantly. > to reproduce run the attached server and client for a short while (30 seconds > on my box). -- This message is automatically generated by JIRA. - You can reply to this email to add a comment to the issue online.
[jira] Commented: (DIRMINA-762) WARN org.apache.mina.core.service.IoProcessor - Create a new selector. Selected is 0, delta = 0
[ https://issues.apache.org/jira/browse/DIRMINA-762?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12831924#action_12831924 ] Emmanuel Lecharny commented on DIRMINA-762: --- Tested with 2.0.0-RC1, no problem so far. Definitively a bug introduced in one of the 47 commits since then ... I'm currently reverting the commits one by one until I see the test running, to find which one introduced the problem. It takes a hell of a time ... > WARN org.apache.mina.core.service.IoProcessor - Create a new selector. > Selected is 0, delta = 0 > > > Key: DIRMINA-762 > URL: https://issues.apache.org/jira/browse/DIRMINA-762 > Project: MINA > Issue Type: Bug > Environment: Linux (2.6.26-2-amd64), java version 1.6.0_12 and also > 1.6.0_18. >Reporter: Omry Yadan >Priority: Critical > Fix For: 2.0.0-RC2 > > Attachments: BufferCodec.java, NettyTestServer.java, > RateCounter.java, Screen shot 2010-02-02 at 7.48.39 PM.png, Screen shot > 2010-02-02 at 7.48.46 PM.png, Screen shot 2010-02-02 at 7.48.59 PM.png, > Screen shot 2010-02-02 at 7.49.13 PM.png, Screen shot 2010-02-02 at 7.49.18 > PM.png, Server.java, StressClient.java > > > Mina server gets into a bad state where it constantly prints : > WARN org.apache.mina.core.service.IoProcessor - Create a new selector. > Selected is 0, delta = 0 > when this happens, server throughput drops significantly. > to reproduce run the attached server and client for a short while (30 seconds > on my box). -- This message is automatically generated by JIRA. - You can reply to this email to add a comment to the issue online.
[jira] Commented: (DIRMINA-762) WARN org.apache.mina.core.service.IoProcessor - Create a new selector. Selected is 0, delta = 0
[ https://issues.apache.org/jira/browse/DIRMINA-762?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12832015#action_12832015 ] Emmanuel Lecharny commented on DIRMINA-762: --- I think that this commit broke the server : http://svn.apache.org/viewvc?view=revision&revision=42 I will check what's inside (I wrote it, so I guess I'm responsible ;) > WARN org.apache.mina.core.service.IoProcessor - Create a new selector. > Selected is 0, delta = 0 > > > Key: DIRMINA-762 > URL: https://issues.apache.org/jira/browse/DIRMINA-762 > Project: MINA > Issue Type: Bug > Environment: Linux (2.6.26-2-amd64), java version 1.6.0_12 and also > 1.6.0_18. >Reporter: Omry Yadan >Priority: Critical > Fix For: 2.0.0-RC2 > > Attachments: BufferCodec.java, NettyTestServer.java, > RateCounter.java, Screen shot 2010-02-02 at 7.48.39 PM.png, Screen shot > 2010-02-02 at 7.48.46 PM.png, Screen shot 2010-02-02 at 7.48.59 PM.png, > Screen shot 2010-02-02 at 7.49.13 PM.png, Screen shot 2010-02-02 at 7.49.18 > PM.png, Server.java, StressClient.java > > > Mina server gets into a bad state where it constantly prints : > WARN org.apache.mina.core.service.IoProcessor - Create a new selector. > Selected is 0, delta = 0 > when this happens, server throughput drops significantly. > to reproduce run the attached server and client for a short while (30 seconds > on my box). -- This message is automatically generated by JIRA. - You can reply to this email to add a comment to the issue online.
[jira] Commented: (DIRMINA-762) WARN org.apache.mina.core.service.IoProcessor - Create a new selector. Selected is 0, delta = 0
[ https://issues.apache.org/jira/browse/DIRMINA-762?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12832023#action_12832023 ] Serge Baranov commented on DIRMINA-762: --- Emmanuel, that's exactly what I found yesterday, you should have started from that commit right away, see http://issues.apache.org/jira/browse/DIRMINA-678?focusedCommentId=12831297&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12831297 . > WARN org.apache.mina.core.service.IoProcessor - Create a new selector. > Selected is 0, delta = 0 > > > Key: DIRMINA-762 > URL: https://issues.apache.org/jira/browse/DIRMINA-762 > Project: MINA > Issue Type: Bug > Environment: Linux (2.6.26-2-amd64), java version 1.6.0_12 and also > 1.6.0_18. >Reporter: Omry Yadan >Priority: Critical > Fix For: 2.0.0-RC2 > > Attachments: BufferCodec.java, NettyTestServer.java, > RateCounter.java, Screen shot 2010-02-02 at 7.48.39 PM.png, Screen shot > 2010-02-02 at 7.48.46 PM.png, Screen shot 2010-02-02 at 7.48.59 PM.png, > Screen shot 2010-02-02 at 7.49.13 PM.png, Screen shot 2010-02-02 at 7.49.18 > PM.png, Server.java, StressClient.java > > > Mina server gets into a bad state where it constantly prints : > WARN org.apache.mina.core.service.IoProcessor - Create a new selector. > Selected is 0, delta = 0 > when this happens, server throughput drops significantly. > to reproduce run the attached server and client for a short while (30 seconds > on my box). -- This message is automatically generated by JIRA. - You can reply to this email to add a comment to the issue online.
[jira] Commented: (DIRMINA-762) WARN org.apache.mina.core.service.IoProcessor - Create a new selector. Selected is 0, delta = 0
[ https://issues.apache.org/jira/browse/DIRMINA-762?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12832029#action_12832029 ] Emmanuel Lecharny commented on DIRMINA-762: --- sh*t... I could have saved me a hell of time... Yeah, Serge, this is THE commit. I'm checking what's inside. It's a pretty long one though... At least, we have a suspect now, time to use appropriate force on him to make him talk :) > WARN org.apache.mina.core.service.IoProcessor - Create a new selector. > Selected is 0, delta = 0 > > > Key: DIRMINA-762 > URL: https://issues.apache.org/jira/browse/DIRMINA-762 > Project: MINA > Issue Type: Bug > Environment: Linux (2.6.26-2-amd64), java version 1.6.0_12 and also > 1.6.0_18. >Reporter: Omry Yadan >Priority: Critical > Fix For: 2.0.0-RC2 > > Attachments: BufferCodec.java, NettyTestServer.java, > RateCounter.java, Screen shot 2010-02-02 at 7.48.39 PM.png, Screen shot > 2010-02-02 at 7.48.46 PM.png, Screen shot 2010-02-02 at 7.48.59 PM.png, > Screen shot 2010-02-02 at 7.49.13 PM.png, Screen shot 2010-02-02 at 7.49.18 > PM.png, Server.java, StressClient.java > > > Mina server gets into a bad state where it constantly prints : > WARN org.apache.mina.core.service.IoProcessor - Create a new selector. > Selected is 0, delta = 0 > when this happens, server throughput drops significantly. > to reproduce run the attached server and client for a short while (30 seconds > on my box). -- This message is automatically generated by JIRA. - You can reply to this email to add a comment to the issue online.
[jira] Commented: (DIRMINA-762) WARN org.apache.mina.core.service.IoProcessor - Create a new selector. Selected is 0, delta = 0
[ https://issues.apache.org/jira/browse/DIRMINA-762?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12832039#action_12832039 ] Emmanuel Lecharny commented on DIRMINA-762: --- Having stripped down all that is irrelevant in this giant patch, I have some interesting parts : - a new atomic flag has been added (wakeupCalled), I don't remember why - this is used to protect the part of the code where we try to detect that the epoll is spinning - I'm wondering if we don't have a race condition here, and stay locked... A quick thread dump could help... > WARN org.apache.mina.core.service.IoProcessor - Create a new selector. > Selected is 0, delta = 0 > > > Key: DIRMINA-762 > URL: https://issues.apache.org/jira/browse/DIRMINA-762 > Project: MINA > Issue Type: Bug > Environment: Linux (2.6.26-2-amd64), java version 1.6.0_12 and also > 1.6.0_18. >Reporter: Omry Yadan >Priority: Critical > Fix For: 2.0.0-RC2 > > Attachments: BufferCodec.java, NettyTestServer.java, > RateCounter.java, Screen shot 2010-02-02 at 7.48.39 PM.png, Screen shot > 2010-02-02 at 7.48.46 PM.png, Screen shot 2010-02-02 at 7.48.59 PM.png, > Screen shot 2010-02-02 at 7.49.13 PM.png, Screen shot 2010-02-02 at 7.49.18 > PM.png, Server.java, StressClient.java > > > Mina server gets into a bad state where it constantly prints : > WARN org.apache.mina.core.service.IoProcessor - Create a new selector. > Selected is 0, delta = 0 > when this happens, server throughput drops significantly. > to reproduce run the attached server and client for a short while (30 seconds > on my box). -- This message is automatically generated by JIRA. - You can reply to this email to add a comment to the issue online.
[jira] Commented: (DIRMINA-762) WARN org.apache.mina.core.service.IoProcessor - Create a new selector. Selected is 0, delta = 0
[ https://issues.apache.org/jira/browse/DIRMINA-762?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12832116#action_12832116 ] Emmanuel Lecharny commented on DIRMINA-762: --- eh, eh, eh, I may have a fix :) The wakeupCalled.getAndSet(false); is not invoked before we do a "continue", so the flag keep the same value ad vitam aeternam. Wrong ... I have the sample running for more than 1 million messages, still at the same rate. Sounds good ! > WARN org.apache.mina.core.service.IoProcessor - Create a new selector. > Selected is 0, delta = 0 > > > Key: DIRMINA-762 > URL: https://issues.apache.org/jira/browse/DIRMINA-762 > Project: MINA > Issue Type: Bug > Environment: Linux (2.6.26-2-amd64), java version 1.6.0_12 and also > 1.6.0_18. >Reporter: Omry Yadan >Priority: Critical > Fix For: 2.0.0-RC2 > > Attachments: BufferCodec.java, NettyTestServer.java, > RateCounter.java, Screen shot 2010-02-02 at 7.48.39 PM.png, Screen shot > 2010-02-02 at 7.48.46 PM.png, Screen shot 2010-02-02 at 7.48.59 PM.png, > Screen shot 2010-02-02 at 7.49.13 PM.png, Screen shot 2010-02-02 at 7.49.18 > PM.png, Server.java, StressClient.java > > > Mina server gets into a bad state where it constantly prints : > WARN org.apache.mina.core.service.IoProcessor - Create a new selector. > Selected is 0, delta = 0 > when this happens, server throughput drops significantly. > to reproduce run the attached server and client for a short while (30 seconds > on my box). -- This message is automatically generated by JIRA. - You can reply to this email to add a comment to the issue online.
[jira] Commented: (DIRMINA-762) WARN org.apache.mina.core.service.IoProcessor - Create a new selector. Selected is 0, delta = 0
[ https://issues.apache.org/jira/browse/DIRMINA-762?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12832140#action_12832140 ] Omry Yadan commented on DIRMINA-762: sounds good :) > WARN org.apache.mina.core.service.IoProcessor - Create a new selector. > Selected is 0, delta = 0 > > > Key: DIRMINA-762 > URL: https://issues.apache.org/jira/browse/DIRMINA-762 > Project: MINA > Issue Type: Bug > Environment: Linux (2.6.26-2-amd64), java version 1.6.0_12 and also > 1.6.0_18. >Reporter: Omry Yadan >Priority: Critical > Fix For: 2.0.0-RC2 > > Attachments: BufferCodec.java, NettyTestServer.java, > RateCounter.java, Screen shot 2010-02-02 at 7.48.39 PM.png, Screen shot > 2010-02-02 at 7.48.46 PM.png, Screen shot 2010-02-02 at 7.48.59 PM.png, > Screen shot 2010-02-02 at 7.49.13 PM.png, Screen shot 2010-02-02 at 7.49.18 > PM.png, Server.java, StressClient.java > > > Mina server gets into a bad state where it constantly prints : > WARN org.apache.mina.core.service.IoProcessor - Create a new selector. > Selected is 0, delta = 0 > when this happens, server throughput drops significantly. > to reproduce run the attached server and client for a short while (30 seconds > on my box). -- This message is automatically generated by JIRA. - You can reply to this email to add a comment to the issue online.
[jira] Commented: (DIRMINA-762) WARN org.apache.mina.core.service.IoProcessor - Create a new selector. Selected is 0, delta = 0
[ https://issues.apache.org/jira/browse/DIRMINA-762?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12833532#action_12833532 ] Omry Yadan commented on DIRMINA-762: the problem I reported appears to be fixed, but this benchmark exposes another performance issue: configure it with 32 long connection, 0 short connection threads and 0 delay between long connection messages: // Number of long lived connections static final int LONG_CONNECTIONS = 32; // Number of threads that spawns short lived connections static final int SHORT_CONNECTIONS_THREADS = 0; // number of ms to sleep before sending a message to the server in the long connection threads static final int LONG_CONNECITON_THREAD_MESSAGE_DELAY = 0; Throughput it very low with this setup (about 10 messages/sec). With the attached Netty test server you I reached 200k-300k messages/sec with this setup. > WARN org.apache.mina.core.service.IoProcessor - Create a new selector. > Selected is 0, delta = 0 > > > Key: DIRMINA-762 > URL: https://issues.apache.org/jira/browse/DIRMINA-762 > Project: MINA > Issue Type: Bug > Environment: Linux (2.6.26-2-amd64), java version 1.6.0_12 and also > 1.6.0_18. >Reporter: Omry Yadan >Priority: Critical > Fix For: 2.0.0-RC2 > > Attachments: BufferCodec.java, NettyTestServer.java, > RateCounter.java, Screen shot 2010-02-02 at 7.48.39 PM.png, Screen shot > 2010-02-02 at 7.48.46 PM.png, Screen shot 2010-02-02 at 7.48.59 PM.png, > Screen shot 2010-02-02 at 7.49.13 PM.png, Screen shot 2010-02-02 at 7.49.18 > PM.png, Server.java, StressClient.java > > > Mina server gets into a bad state where it constantly prints : > WARN org.apache.mina.core.service.IoProcessor - Create a new selector. > Selected is 0, delta = 0 > when this happens, server throughput drops significantly. > to reproduce run the attached server and client for a short while (30 seconds > on my box). -- This message is automatically generated by JIRA. - You can reply to this email to add a comment to the issue online.
[jira] Commented: (DIRMINA-762) WARN org.apache.mina.core.service.IoProcessor - Create a new selector. Selected is 0, delta = 0
[ https://issues.apache.org/jira/browse/DIRMINA-762?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12833798#action_12833798 ] Emmanuel Lecharny commented on DIRMINA-762: --- I moved the issue on https://issues.apache.org/jira/browse/DIRMINA-764 > WARN org.apache.mina.core.service.IoProcessor - Create a new selector. > Selected is 0, delta = 0 > > > Key: DIRMINA-762 > URL: https://issues.apache.org/jira/browse/DIRMINA-762 > Project: MINA > Issue Type: Bug > Environment: Linux (2.6.26-2-amd64), java version 1.6.0_12 and also > 1.6.0_18. >Reporter: Omry Yadan >Priority: Critical > Fix For: 2.0.0-RC2 > > Attachments: BufferCodec.java, NettyTestServer.java, > RateCounter.java, Screen shot 2010-02-02 at 7.48.39 PM.png, Screen shot > 2010-02-02 at 7.48.46 PM.png, Screen shot 2010-02-02 at 7.48.59 PM.png, > Screen shot 2010-02-02 at 7.49.13 PM.png, Screen shot 2010-02-02 at 7.49.18 > PM.png, Server.java, StressClient.java > > > Mina server gets into a bad state where it constantly prints : > WARN org.apache.mina.core.service.IoProcessor - Create a new selector. > Selected is 0, delta = 0 > when this happens, server throughput drops significantly. > to reproduce run the attached server and client for a short while (30 seconds > on my box). -- This message is automatically generated by JIRA. - You can reply to this email to add a comment to the issue online.
Re: [jira] Commented: (DIRMINA-762) WARN org.apache.mina.core.service.IoProcessor - Create a new selector. Selected is 0, delta = 0
Hi All, Sorry for this email which will undoubtedly cause frustration. Can somebody please remove me from the list? The process for unsubscribing fails every time. Great technology by the way! - Original Message - From: "Emmanuel Lecharny (JIRA)" To: dev@mina.apache.org Sent: Wednesday, 10 February, 2010 16:03:28 GMT +00:00 Monrovia Subject: [jira] Commented: (DIRMINA-762) WARN org.apache.mina.core.service.IoProcessor - Create a new selector. Selected is 0, delta = 0 [ https://issues.apache.org/jira/browse/DIRMINA-762?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12832039#action_12832039 ] Emmanuel Lecharny commented on DIRMINA-762: --- Having stripped down all that is irrelevant in this giant patch, I have some interesting parts : - a new atomic flag has been added (wakeupCalled), I don't remember why - this is used to protect the part of the code where we try to detect that the epoll is spinning - I'm wondering if we don't have a race condition here, and stay locked... A quick thread dump could help... > WARN org.apache.mina.core.service.IoProcessor - Create a new selector. > Selected is 0, delta = 0 > > > > Key: DIRMINA-762 > URL: https://issues.apache.org/jira/browse/DIRMINA-762 > Project: MINA > Issue Type: Bug > Environment: Linux (2.6.26-2-amd64), java version 1.6.0_12 and also 1.6.0_18. > Reporter: Omry Yadan > Priority: Critical > Fix For: 2.0.0-RC2 > > Attachments: BufferCodec.java, NettyTestServer.java, RateCounter.java, Screen > shot 2010-02-02 at 7.48.39 PM.png, Screen shot 2010-02-02 at 7.48.46 PM.png, > Screen shot 2010-02-02 at 7.48.59 PM.png, Screen shot 2010-02-02 at 7.49.13 > PM.png, Screen shot 2010-02-02 at 7.49.18 PM.png, Server.java, > StressClient.java > > > Mina server gets into a bad state where it constantly prints : > WARN org.apache.mina.core.service.IoProcessor - Create a new selector. > Selected is 0, delta = 0 > when this happens, server throughput drops significantly. > to reproduce run the attached server and client for a short while (30 seconds > on my box). -- This message is automatically generated by JIRA. - You can reply to this email to add a comment to the issue online.
Re: [jira] Commented: (DIRMINA-762) WARN org.apache.mina.core.service.IoProcessor - Create a new selector. Selected is 0, delta = 0
On 2/10/10 5:05 PM, Jon Folland, Nativ Ltd wrote: Hi All, Hi, Sorry for this email which will undoubtedly cause frustration. Can somebody please remove me from the list? The process for unsubscribing fails every time. This was a gentle request, so I unsubscribed you :) If only everyone using mailing lists were as polite as you are ... If you still receive mails from this mailing list, feel free to contact me. -- Regards, Cordialement, Emmanuel Lécharny www.nextury.com
