You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@mina.apache.org by "Mauritz Lovgren (JIRA)" <ji...@apache.org> on 2009/04/01 13:01:13 UTC
[jira] Created: (DIRMINA-681) Strange CPU peak occuring at fixed
interval when several thousand connections active
Strange CPU peak occuring at fixed interval when several thousand connections active
------------------------------------------------------------------------------------
Key: DIRMINA-681
URL: https://issues.apache.org/jira/browse/DIRMINA-681
Project: MINA
Issue Type: Bug
Components: Core
Affects Versions: 2.0.0-M4
Environment: Windows Vista Ultimate 64-bit (on 64-bit Sun JDK 1.6.0_12)
Reporter: Mauritz Lovgren
Observing strange CPU activity occuring at regular (seemingly fixed) interval with no protocol traffic activity.
See attached window capture of task manager that shows this with 3000 active connections.
Is there some kind of cleanup occuring within MINA core at a predefined interval?
The 3000 connections in the example above connects within 250 seconds. A normal situation would be that these connections are established over a longer period of time, perhaps spreading the CPU peaks shown above as well, flattening the curve.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (DIRMINA-681) Strange CPU peak occuring at fixed
interval when several thousand connections active
Posted by "Mauritz Lovgren (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/DIRMINA-681?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12830949#action_12830949 ]
Mauritz Lovgren commented on DIRMINA-681:
-----------------------------------------
Would it be possible to perform an isolated port of the epoll fix on top of RC1 to be able to verify it without having to care about other regression issues in RC2?
> Strange CPU peak occuring at fixed interval when several thousand connections active
> ------------------------------------------------------------------------------------
>
> Key: DIRMINA-681
> URL: https://issues.apache.org/jira/browse/DIRMINA-681
> Project: MINA
> Issue Type: Task
> Components: Core
> Affects Versions: 2.0.0-M4, 2.0.0-RC1
> Environment: Windows Vista Ultimate 64-bit (on 64-bit Sun JDK 1.6.0_18). Intel Core 2 Quad Core Q9300 2,5 GHz, 8 GB RAM
> Reporter: Mauritz Lovgren
> Fix For: 2.0.0
>
> Attachments: screenshot-1.jpg, screenshot-2.jpg, screenshot-3.jpg, screenshot-4.jpg
>
>
> Observing strange CPU activity occuring at regular (seemingly fixed) interval with no protocol traffic activity.
> See attached window capture of task manager that shows this with 3000 active connections.
> Is there some kind of cleanup occuring within MINA core at a predefined interval?
> The 3000 connections in the example above connects within 250 seconds. A normal situation would be that these connections are established over a longer period of time, perhaps spreading the CPU peaks shown above as well, flattening the curve.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (DIRMINA-681) Strange CPU peak occuring at fixed
interval when several thousand connections active
Posted by "Mauritz Lovgren (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/DIRMINA-681?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12694476#action_12694476 ]
Mauritz Lovgren commented on DIRMINA-681:
-----------------------------------------
The CPU peak does not (seem) to affect the performance noticeably (at least I have not been bothered by it). The CPU spike seems to increase linearly with number of connections though, so I fear that it might start causing problems when number of connections get very high (10000+).
Updating environment with hardware spec.
> Strange CPU peak occuring at fixed interval when several thousand connections active
> ------------------------------------------------------------------------------------
>
> Key: DIRMINA-681
> URL: https://issues.apache.org/jira/browse/DIRMINA-681
> Project: MINA
> Issue Type: Bug
> Components: Core
> Affects Versions: 2.0.0-M4
> Environment: Windows Vista Ultimate 64-bit (on 64-bit Sun JDK 1.6.0_12)
> Reporter: Mauritz Lovgren
> Attachments: screenshot-1.jpg, screenshot-2.jpg, screenshot-3.jpg
>
>
> Observing strange CPU activity occuring at regular (seemingly fixed) interval with no protocol traffic activity.
> See attached window capture of task manager that shows this with 3000 active connections.
> Is there some kind of cleanup occuring within MINA core at a predefined interval?
> The 3000 connections in the example above connects within 250 seconds. A normal situation would be that these connections are established over a longer period of time, perhaps spreading the CPU peaks shown above as well, flattening the curve.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (DIRMINA-681) Strange CPU peak occuring at fixed
interval when several thousand connections active
Posted by "Mauritz Lovgren (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/DIRMINA-681?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Mauritz Lovgren updated DIRMINA-681:
------------------------------------
Attachment: screenshot-3.jpg
Here, we see that no extra message activity is occuring during the CPU peak time frame.
> Strange CPU peak occuring at fixed interval when several thousand connections active
> ------------------------------------------------------------------------------------
>
> Key: DIRMINA-681
> URL: https://issues.apache.org/jira/browse/DIRMINA-681
> Project: MINA
> Issue Type: Bug
> Components: Core
> Affects Versions: 2.0.0-M4
> Environment: Windows Vista Ultimate 64-bit (on 64-bit Sun JDK 1.6.0_12)
> Reporter: Mauritz Lovgren
> Attachments: screenshot-1.jpg, screenshot-2.jpg, screenshot-3.jpg
>
>
> Observing strange CPU activity occuring at regular (seemingly fixed) interval with no protocol traffic activity.
> See attached window capture of task manager that shows this with 3000 active connections.
> Is there some kind of cleanup occuring within MINA core at a predefined interval?
> The 3000 connections in the example above connects within 250 seconds. A normal situation would be that these connections are established over a longer period of time, perhaps spreading the CPU peaks shown above as well, flattening the curve.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (DIRMINA-681) Strange CPU peak occuring at fixed
interval when several thousand connections active
Posted by "Emmanuel Lecharny (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/DIRMINA-681?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12830956#action_12830956 ]
Emmanuel Lecharny commented on DIRMINA-681:
-------------------------------------------
Sure.
let's do that. I will try to provide a patch later today.
> Strange CPU peak occuring at fixed interval when several thousand connections active
> ------------------------------------------------------------------------------------
>
> Key: DIRMINA-681
> URL: https://issues.apache.org/jira/browse/DIRMINA-681
> Project: MINA
> Issue Type: Task
> Components: Core
> Affects Versions: 2.0.0-M4, 2.0.0-RC1
> Environment: Windows Vista Ultimate 64-bit (on 64-bit Sun JDK 1.6.0_18). Intel Core 2 Quad Core Q9300 2,5 GHz, 8 GB RAM
> Reporter: Mauritz Lovgren
> Fix For: 2.0.0
>
> Attachments: screenshot-1.jpg, screenshot-2.jpg, screenshot-3.jpg, screenshot-4.jpg
>
>
> Observing strange CPU activity occuring at regular (seemingly fixed) interval with no protocol traffic activity.
> See attached window capture of task manager that shows this with 3000 active connections.
> Is there some kind of cleanup occuring within MINA core at a predefined interval?
> The 3000 connections in the example above connects within 250 seconds. A normal situation would be that these connections are established over a longer period of time, perhaps spreading the CPU peaks shown above as well, flattening the curve.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (DIRMINA-681) Strange CPU peak occuring at fixed
interval when several thousand connections active
Posted by "Mauritz Lovgren (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/DIRMINA-681?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12805003#action_12805003 ]
Mauritz Lovgren commented on DIRMINA-681:
-----------------------------------------
Hm. When running with trunk (RC2-SNAPSHOT of today), I am not able to get more than 5-600 client connections before the proxy stops responding and gets no network IO... Something changed between RC1 and RC2 that will cause this (message handling etc etc?)...
> Strange CPU peak occuring at fixed interval when several thousand connections active
> ------------------------------------------------------------------------------------
>
> Key: DIRMINA-681
> URL: https://issues.apache.org/jira/browse/DIRMINA-681
> Project: MINA
> Issue Type: Task
> Components: Core
> Affects Versions: 2.0.0-M4, 2.0.0-RC1
> Environment: Windows Vista Ultimate 64-bit (on 64-bit Sun JDK 1.6.0_18). Intel Core 2 Quad Core Q9300 2,5 GHz, 8 GB RAM
> Reporter: Mauritz Lovgren
> Fix For: 2.0.0
>
> Attachments: screenshot-1.jpg, screenshot-2.jpg, screenshot-3.jpg, screenshot-4.jpg
>
>
> Observing strange CPU activity occuring at regular (seemingly fixed) interval with no protocol traffic activity.
> See attached window capture of task manager that shows this with 3000 active connections.
> Is there some kind of cleanup occuring within MINA core at a predefined interval?
> The 3000 connections in the example above connects within 250 seconds. A normal situation would be that these connections are established over a longer period of time, perhaps spreading the CPU peaks shown above as well, flattening the curve.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (DIRMINA-681) Strange CPU peak occuring at fixed
interval when several thousand connections active
Posted by "Mauritz Lovgren (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/DIRMINA-681?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Mauritz Lovgren updated DIRMINA-681:
------------------------------------
Attachment: screenshot-2.jpg
JConsole output for the process when the CPU peak is occuring. No extra garbage collection found to occur during the CPU peak.
> Strange CPU peak occuring at fixed interval when several thousand connections active
> ------------------------------------------------------------------------------------
>
> Key: DIRMINA-681
> URL: https://issues.apache.org/jira/browse/DIRMINA-681
> Project: MINA
> Issue Type: Bug
> Components: Core
> Affects Versions: 2.0.0-M4
> Environment: Windows Vista Ultimate 64-bit (on 64-bit Sun JDK 1.6.0_12)
> Reporter: Mauritz Lovgren
> Attachments: screenshot-1.jpg, screenshot-2.jpg, screenshot-3.jpg
>
>
> Observing strange CPU activity occuring at regular (seemingly fixed) interval with no protocol traffic activity.
> See attached window capture of task manager that shows this with 3000 active connections.
> Is there some kind of cleanup occuring within MINA core at a predefined interval?
> The 3000 connections in the example above connects within 250 seconds. A normal situation would be that these connections are established over a longer period of time, perhaps spreading the CPU peaks shown above as well, flattening the curve.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (DIRMINA-681) Strange CPU peak occuring at fixed
interval when several thousand connections active
Posted by "Mauritz Lovgren (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/DIRMINA-681?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12804433#action_12804433 ]
Mauritz Lovgren commented on DIRMINA-681:
-----------------------------------------
I will perform profiling this week to identify the source of the CPU peaks...
> Strange CPU peak occuring at fixed interval when several thousand connections active
> ------------------------------------------------------------------------------------
>
> Key: DIRMINA-681
> URL: https://issues.apache.org/jira/browse/DIRMINA-681
> Project: MINA
> Issue Type: Task
> Components: Core
> Affects Versions: 2.0.0-M4
> Environment: Windows Vista Ultimate 64-bit (on 64-bit Sun JDK 1.6.0_12). Intel Core 2 Quad Core Q9300 2,5 GHz, 8 GB RAM
> Reporter: Mauritz Lovgren
> Fix For: 2.0.0
>
> Attachments: screenshot-1.jpg, screenshot-2.jpg, screenshot-3.jpg
>
>
> Observing strange CPU activity occuring at regular (seemingly fixed) interval with no protocol traffic activity.
> See attached window capture of task manager that shows this with 3000 active connections.
> Is there some kind of cleanup occuring within MINA core at a predefined interval?
> The 3000 connections in the example above connects within 250 seconds. A normal situation would be that these connections are established over a longer period of time, perhaps spreading the CPU peaks shown above as well, flattening the curve.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (DIRMINA-681) Strange CPU peak occuring at fixed
interval when several thousand connections active
Posted by "Mauritz Lovgren (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/DIRMINA-681?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12694479#action_12694479 ]
Mauritz Lovgren commented on DIRMINA-681:
-----------------------------------------
This is easily reproducable by creating a test server having thousands of clients doing almost nothing (just sending a heartbeat or similar every minute).
> Strange CPU peak occuring at fixed interval when several thousand connections active
> ------------------------------------------------------------------------------------
>
> Key: DIRMINA-681
> URL: https://issues.apache.org/jira/browse/DIRMINA-681
> Project: MINA
> Issue Type: Bug
> Components: Core
> Affects Versions: 2.0.0-M4
> Environment: Windows Vista Ultimate 64-bit (on 64-bit Sun JDK 1.6.0_12). Intel Core 2 Quad Core Q9300 2,5 GHz, 8 GB RAM
> Reporter: Mauritz Lovgren
> Attachments: screenshot-1.jpg, screenshot-2.jpg, screenshot-3.jpg
>
>
> Observing strange CPU activity occuring at regular (seemingly fixed) interval with no protocol traffic activity.
> See attached window capture of task manager that shows this with 3000 active connections.
> Is there some kind of cleanup occuring within MINA core at a predefined interval?
> The 3000 connections in the example above connects within 250 seconds. A normal situation would be that these connections are established over a longer period of time, perhaps spreading the CPU peaks shown above as well, flattening the curve.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (DIRMINA-681) Strange CPU peak occuring at fixed
interval when several thousand connections active
Posted by "Mauritz Lovgren (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/DIRMINA-681?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12804555#action_12804555 ]
Mauritz Lovgren commented on DIRMINA-681:
-----------------------------------------
I will do it the first thing in the morning :-).
> Strange CPU peak occuring at fixed interval when several thousand connections active
> ------------------------------------------------------------------------------------
>
> Key: DIRMINA-681
> URL: https://issues.apache.org/jira/browse/DIRMINA-681
> Project: MINA
> Issue Type: Task
> Components: Core
> Affects Versions: 2.0.0-M4, 2.0.0-RC1
> Environment: Windows Vista Ultimate 64-bit (on 64-bit Sun JDK 1.6.0_18). Intel Core 2 Quad Core Q9300 2,5 GHz, 8 GB RAM
> Reporter: Mauritz Lovgren
> Fix For: 2.0.0
>
> Attachments: screenshot-1.jpg, screenshot-2.jpg, screenshot-3.jpg, screenshot-4.jpg
>
>
> Observing strange CPU activity occuring at regular (seemingly fixed) interval with no protocol traffic activity.
> See attached window capture of task manager that shows this with 3000 active connections.
> Is there some kind of cleanup occuring within MINA core at a predefined interval?
> The 3000 connections in the example above connects within 250 seconds. A normal situation would be that these connections are established over a longer period of time, perhaps spreading the CPU peaks shown above as well, flattening the curve.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (DIRMINA-681) Strange CPU peak occuring at fixed
interval when several thousand connections active
Posted by "Victor N (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/DIRMINA-681?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12694648#action_12694648 ]
Victor N commented on DIRMINA-681:
----------------------------------
Mauritz, yes, I see that there is no GC activity on your diagrams.
It's difficult to say what may be wrong; I would propose you to attach to you test server with a profiler tool and make CPU profiling - you will see which methods consume most of CPU time. Usually, I use YourKit - http://www.yourkit.com/download/
> Strange CPU peak occuring at fixed interval when several thousand connections active
> ------------------------------------------------------------------------------------
>
> Key: DIRMINA-681
> URL: https://issues.apache.org/jira/browse/DIRMINA-681
> Project: MINA
> Issue Type: Bug
> Components: Core
> Affects Versions: 2.0.0-M4
> Environment: Windows Vista Ultimate 64-bit (on 64-bit Sun JDK 1.6.0_12). Intel Core 2 Quad Core Q9300 2,5 GHz, 8 GB RAM
> Reporter: Mauritz Lovgren
> Attachments: screenshot-1.jpg, screenshot-2.jpg, screenshot-3.jpg
>
>
> Observing strange CPU activity occuring at regular (seemingly fixed) interval with no protocol traffic activity.
> See attached window capture of task manager that shows this with 3000 active connections.
> Is there some kind of cleanup occuring within MINA core at a predefined interval?
> The 3000 connections in the example above connects within 250 seconds. A normal situation would be that these connections are established over a longer period of time, perhaps spreading the CPU peaks shown above as well, flattening the curve.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (DIRMINA-681) Strange CPU peak occuring at fixed
interval when several thousand connections active
Posted by "Mauritz Lovgren (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/DIRMINA-681?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Mauritz Lovgren updated DIRMINA-681:
------------------------------------
Attachment: screenshot-1.jpg
Task manager capture that shows the strange CPU peaks caused by MINA having 3000 client connections active with no traffic.
> Strange CPU peak occuring at fixed interval when several thousand connections active
> ------------------------------------------------------------------------------------
>
> Key: DIRMINA-681
> URL: https://issues.apache.org/jira/browse/DIRMINA-681
> Project: MINA
> Issue Type: Bug
> Components: Core
> Affects Versions: 2.0.0-M4
> Environment: Windows Vista Ultimate 64-bit (on 64-bit Sun JDK 1.6.0_12)
> Reporter: Mauritz Lovgren
> Attachments: screenshot-1.jpg
>
>
> Observing strange CPU activity occuring at regular (seemingly fixed) interval with no protocol traffic activity.
> See attached window capture of task manager that shows this with 3000 active connections.
> Is there some kind of cleanup occuring within MINA core at a predefined interval?
> The 3000 connections in the example above connects within 250 seconds. A normal situation would be that these connections are established over a longer period of time, perhaps spreading the CPU peaks shown above as well, flattening the curve.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (DIRMINA-681) Strange CPU peak occuring at fixed
interval when several thousand connections active
Posted by "Mauritz Lovgren (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/DIRMINA-681?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12833777#action_12833777 ]
Mauritz Lovgren commented on DIRMINA-681:
-----------------------------------------
I performed a new test run from trunk today, and after the last few check-ins (revisions after 27.jan.2010), the performance problem seems to be gone.
The CPU peaks are still present, though. Did you simply revert the epoll fix in trunk?
> Strange CPU peak occuring at fixed interval when several thousand connections active
> ------------------------------------------------------------------------------------
>
> Key: DIRMINA-681
> URL: https://issues.apache.org/jira/browse/DIRMINA-681
> Project: MINA
> Issue Type: Task
> Components: Core
> Affects Versions: 2.0.0-M4, 2.0.0-RC1
> Environment: Windows Vista Ultimate 64-bit (on 64-bit Sun JDK 1.6.0_18). Intel Core 2 Quad Core Q9300 2,5 GHz, 8 GB RAM
> Reporter: Mauritz Lovgren
> Fix For: 2.0.0
>
> Attachments: screenshot-1.jpg, screenshot-2.jpg, screenshot-3.jpg, screenshot-4.jpg
>
>
> Observing strange CPU activity occuring at regular (seemingly fixed) interval with no protocol traffic activity.
> See attached window capture of task manager that shows this with 3000 active connections.
> Is there some kind of cleanup occuring within MINA core at a predefined interval?
> The 3000 connections in the example above connects within 250 seconds. A normal situation would be that these connections are established over a longer period of time, perhaps spreading the CPU peaks shown above as well, flattening the curve.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (DIRMINA-681) Strange CPU peak occuring at fixed
interval when several thousand connections active
Posted by "Mauritz Lovgren (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/DIRMINA-681?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12834180#action_12834180 ]
Mauritz Lovgren commented on DIRMINA-681:
-----------------------------------------
Ran a new loadtest today (on trunk RC2 revision 909546).
I have removed all home-grown executor filters etc etc to run NioAcceptor with standard CORE mina defaults only.
8500 clients connecting to one NioAcceptor configured to run using 1 thread (new NioSocketAcceptor(1)) just to keep things simple.
The clients are all running on a separate machine from the server. 1Gb LAN, but very small messages and low bandwidth needed for test.
The clients connect and thereafter send a simple IsAlive message upon idle timeout (120 seconds) and stay connected forever.
The clients are connecting at a rate of about 10 clients per second.
My observations are as follows:
- At startup everyting looks ok with only two MINA threads running, the NioAcceptor and one NioProcessor.
- As clients log on, the WindowSelectorImpl somehow recons that it needs to start SelectorThread instances to cope with traffic (?).
After about each 1000 connections, a new selector thread is created.
- The WindowSelectorImpl$SelectorThread instances are linearly causing CPU overhead after they are created. Before they exist, no extra CPU is seen.
- The profiler claims that the bit of code that is causing the CPU overhead is within the sun.nio.ch.WindowSelectorImpl$SubSelector.poll0() method which is called by each SelectorThread.run().
Could MINA somehow be (mis)using the select() in a way that causes this situation to occur? (E.g. calling it too often, or not handling the select keys properly etc etc).
And, I wonder if this might be the epoll bug at all?
> Strange CPU peak occuring at fixed interval when several thousand connections active
> ------------------------------------------------------------------------------------
>
> Key: DIRMINA-681
> URL: https://issues.apache.org/jira/browse/DIRMINA-681
> Project: MINA
> Issue Type: Task
> Components: Core
> Affects Versions: 2.0.0-M4, 2.0.0-RC1
> Environment: Windows Vista Ultimate 64-bit (on 64-bit Sun JDK 1.6.0_18). Intel Core 2 Quad Core Q9300 2,5 GHz, 8 GB RAM
> Reporter: Mauritz Lovgren
> Fix For: 2.0.0
>
> Attachments: screenshot-1.jpg, screenshot-2.jpg, screenshot-3.jpg, screenshot-4.jpg
>
>
> Observing strange CPU activity occuring at regular (seemingly fixed) interval with no protocol traffic activity.
> See attached window capture of task manager that shows this with 3000 active connections.
> Is there some kind of cleanup occuring within MINA core at a predefined interval?
> The 3000 connections in the example above connects within 250 seconds. A normal situation would be that these connections are established over a longer period of time, perhaps spreading the CPU peaks shown above as well, flattening the curve.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (DIRMINA-681) Strange CPU peak occuring at fixed
interval when several thousand connections active
Posted by "Emmanuel Lecharny (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/DIRMINA-681?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12804544#action_12804544 ]
Emmanuel Lecharny commented on DIRMINA-681:
-------------------------------------------
What would be cool is to conduct a bench with the trunk version : it supposed to contain a fix for this problem (hopefully)
> Strange CPU peak occuring at fixed interval when several thousand connections active
> ------------------------------------------------------------------------------------
>
> Key: DIRMINA-681
> URL: https://issues.apache.org/jira/browse/DIRMINA-681
> Project: MINA
> Issue Type: Task
> Components: Core
> Affects Versions: 2.0.0-M4, 2.0.0-RC1
> Environment: Windows Vista Ultimate 64-bit (on 64-bit Sun JDK 1.6.0_18). Intel Core 2 Quad Core Q9300 2,5 GHz, 8 GB RAM
> Reporter: Mauritz Lovgren
> Fix For: 2.0.0
>
> Attachments: screenshot-1.jpg, screenshot-2.jpg, screenshot-3.jpg, screenshot-4.jpg
>
>
> Observing strange CPU activity occuring at regular (seemingly fixed) interval with no protocol traffic activity.
> See attached window capture of task manager that shows this with 3000 active connections.
> Is there some kind of cleanup occuring within MINA core at a predefined interval?
> The 3000 connections in the example above connects within 250 seconds. A normal situation would be that these connections are established over a longer period of time, perhaps spreading the CPU peaks shown above as well, flattening the curve.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (DIRMINA-681) Strange CPU peak occuring at fixed
interval when several thousand connections active
Posted by "Mauritz Lovgren (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/DIRMINA-681?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12839493#action_12839493 ]
Mauritz Lovgren commented on DIRMINA-681:
-----------------------------------------
The changes in trunk over the previous weeks have caused different performance results during my tests. As of the week-old revisions, the load tests halt after few minutes of traffic and CPU rise to 100% making the entire host unresponsive. I noticed more fixes on the Nio processing code late last week, but I have not had the time to test further. Going for a new run early this week :-). Porting my communications framework to support the latest Netty version as weel to see if it exhibits the same performance problems.
> Strange CPU peak occuring at fixed interval when several thousand connections active
> ------------------------------------------------------------------------------------
>
> Key: DIRMINA-681
> URL: https://issues.apache.org/jira/browse/DIRMINA-681
> Project: MINA
> Issue Type: Task
> Components: Core
> Affects Versions: 2.0.0-M4, 2.0.0-RC1
> Environment: Windows Vista Ultimate 64-bit (on 64-bit Sun JDK 1.6.0_18). Intel Core 2 Quad Core Q9300 2,5 GHz, 8 GB RAM
> Reporter: Mauritz Lovgren
> Fix For: 2.0.0
>
> Attachments: screenshot-1.jpg, screenshot-2.jpg, screenshot-3.jpg, screenshot-4.jpg
>
>
> Observing strange CPU activity occuring at regular (seemingly fixed) interval with no protocol traffic activity.
> See attached window capture of task manager that shows this with 3000 active connections.
> Is there some kind of cleanup occuring within MINA core at a predefined interval?
> The 3000 connections in the example above connects within 250 seconds. A normal situation would be that these connections are established over a longer period of time, perhaps spreading the CPU peaks shown above as well, flattening the curve.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (DIRMINA-681) Strange CPU peak occuring at fixed
interval when several thousand connections active
Posted by "Emmanuel Lecharny (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/DIRMINA-681?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12805688#action_12805688 ]
Emmanuel Lecharny commented on DIRMINA-681:
-------------------------------------------
There was definitively a bug : I forgot to re-attach the session to the selectionKey when I registered it to the newly created selector.
http://svn.apache.org/viewvc?rev=903897&view=rev
Can you tests it o trunk again ?
> Strange CPU peak occuring at fixed interval when several thousand connections active
> ------------------------------------------------------------------------------------
>
> Key: DIRMINA-681
> URL: https://issues.apache.org/jira/browse/DIRMINA-681
> Project: MINA
> Issue Type: Task
> Components: Core
> Affects Versions: 2.0.0-M4, 2.0.0-RC1
> Environment: Windows Vista Ultimate 64-bit (on 64-bit Sun JDK 1.6.0_18). Intel Core 2 Quad Core Q9300 2,5 GHz, 8 GB RAM
> Reporter: Mauritz Lovgren
> Fix For: 2.0.0
>
> Attachments: screenshot-1.jpg, screenshot-2.jpg, screenshot-3.jpg, screenshot-4.jpg
>
>
> Observing strange CPU activity occuring at regular (seemingly fixed) interval with no protocol traffic activity.
> See attached window capture of task manager that shows this with 3000 active connections.
> Is there some kind of cleanup occuring within MINA core at a predefined interval?
> The 3000 connections in the example above connects within 250 seconds. A normal situation would be that these connections are established over a longer period of time, perhaps spreading the CPU peaks shown above as well, flattening the curve.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Issue Comment Edited: (DIRMINA-681) Strange CPU peak
occuring at fixed interval when several thousand connections active
Posted by "Mauritz Lovgren (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/DIRMINA-681?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12694479#action_12694479 ]
Mauritz Lovgren edited comment on DIRMINA-681 at 4/1/09 4:31 AM:
-----------------------------------------------------------------
This is easily reproducable by creating a test server having thousands of clients doing almost nothing (just sending a heartbeat or similar every minute). The CPU peak seems to occur every 10 - 12 minutes.
was (Author: mauritz):
This is easily reproducable by creating a test server having thousands of clients doing almost nothing (just sending a heartbeat or similar every minute).
> Strange CPU peak occuring at fixed interval when several thousand connections active
> ------------------------------------------------------------------------------------
>
> Key: DIRMINA-681
> URL: https://issues.apache.org/jira/browse/DIRMINA-681
> Project: MINA
> Issue Type: Bug
> Components: Core
> Affects Versions: 2.0.0-M4
> Environment: Windows Vista Ultimate 64-bit (on 64-bit Sun JDK 1.6.0_12). Intel Core 2 Quad Core Q9300 2,5 GHz, 8 GB RAM
> Reporter: Mauritz Lovgren
> Attachments: screenshot-1.jpg, screenshot-2.jpg, screenshot-3.jpg
>
>
> Observing strange CPU activity occuring at regular (seemingly fixed) interval with no protocol traffic activity.
> See attached window capture of task manager that shows this with 3000 active connections.
> Is there some kind of cleanup occuring within MINA core at a predefined interval?
> The 3000 connections in the example above connects within 250 seconds. A normal situation would be that these connections are established over a longer period of time, perhaps spreading the CPU peaks shown above as well, flattening the curve.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (DIRMINA-681) Strange CPU peak occuring at fixed
interval when several thousand connections active
Posted by "Victor N (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/DIRMINA-681?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12694470#action_12694470 ]
Victor N commented on DIRMINA-681:
----------------------------------
Can you attach a picture from JConsole - CPU usage and heap usage at the same time? Maybe the problem is an intensive garbage collection and not mina itself?
> Strange CPU peak occuring at fixed interval when several thousand connections active
> ------------------------------------------------------------------------------------
>
> Key: DIRMINA-681
> URL: https://issues.apache.org/jira/browse/DIRMINA-681
> Project: MINA
> Issue Type: Bug
> Components: Core
> Affects Versions: 2.0.0-M4
> Environment: Windows Vista Ultimate 64-bit (on 64-bit Sun JDK 1.6.0_12)
> Reporter: Mauritz Lovgren
> Attachments: screenshot-1.jpg
>
>
> Observing strange CPU activity occuring at regular (seemingly fixed) interval with no protocol traffic activity.
> See attached window capture of task manager that shows this with 3000 active connections.
> Is there some kind of cleanup occuring within MINA core at a predefined interval?
> The 3000 connections in the example above connects within 250 seconds. A normal situation would be that these connections are established over a longer period of time, perhaps spreading the CPU peaks shown above as well, flattening the curve.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (DIRMINA-681) Strange CPU peak occuring at fixed
interval when several thousand connections active
Posted by "Emmanuel Lecharny (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/DIRMINA-681?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Emmanuel Lecharny updated DIRMINA-681:
--------------------------------------
Fix Version/s: 2.0.0
Issue Type: Task (was: Bug)
I don't think that's a bug, so I switched the status to Task, and move it to 2.0.0.
if someone can do some analysis on this, with some thread dump, we can go father. Right now, I would say that we should just wait and see.
> Strange CPU peak occuring at fixed interval when several thousand connections active
> ------------------------------------------------------------------------------------
>
> Key: DIRMINA-681
> URL: https://issues.apache.org/jira/browse/DIRMINA-681
> Project: MINA
> Issue Type: Task
> Components: Core
> Affects Versions: 2.0.0-M4
> Environment: Windows Vista Ultimate 64-bit (on 64-bit Sun JDK 1.6.0_12). Intel Core 2 Quad Core Q9300 2,5 GHz, 8 GB RAM
> Reporter: Mauritz Lovgren
> Fix For: 2.0.0
>
> Attachments: screenshot-1.jpg, screenshot-2.jpg, screenshot-3.jpg
>
>
> Observing strange CPU activity occuring at regular (seemingly fixed) interval with no protocol traffic activity.
> See attached window capture of task manager that shows this with 3000 active connections.
> Is there some kind of cleanup occuring within MINA core at a predefined interval?
> The 3000 connections in the example above connects within 250 seconds. A normal situation would be that these connections are established over a longer period of time, perhaps spreading the CPU peaks shown above as well, flattening the curve.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (DIRMINA-681) Strange CPU peak occuring at fixed
interval when several thousand connections active
Posted by "Emmanuel Lecharny (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/DIRMINA-681?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12804535#action_12804535 ]
Emmanuel Lecharny commented on DIRMINA-681:
-------------------------------------------
The epoll bug is *not* solved with jdk 1.6.0_18. :/
Have you profiled with trunk ?
> Strange CPU peak occuring at fixed interval when several thousand connections active
> ------------------------------------------------------------------------------------
>
> Key: DIRMINA-681
> URL: https://issues.apache.org/jira/browse/DIRMINA-681
> Project: MINA
> Issue Type: Task
> Components: Core
> Affects Versions: 2.0.0-M4
> Environment: Windows Vista Ultimate 64-bit (on 64-bit Sun JDK 1.6.0_18). Intel Core 2 Quad Core Q9300 2,5 GHz, 8 GB RAM
> Reporter: Mauritz Lovgren
> Fix For: 2.0.0
>
> Attachments: screenshot-1.jpg, screenshot-2.jpg, screenshot-3.jpg, screenshot-4.jpg
>
>
> Observing strange CPU activity occuring at regular (seemingly fixed) interval with no protocol traffic activity.
> See attached window capture of task manager that shows this with 3000 active connections.
> Is there some kind of cleanup occuring within MINA core at a predefined interval?
> The 3000 connections in the example above connects within 250 seconds. A normal situation would be that these connections are established over a longer period of time, perhaps spreading the CPU peaks shown above as well, flattening the curve.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (DIRMINA-681) Strange CPU peak occuring at fixed
interval when several thousand connections active
Posted by "Mauritz Lovgren (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/DIRMINA-681?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12841312#action_12841312 ]
Mauritz Lovgren commented on DIRMINA-681:
-----------------------------------------
Having re-written our network framework to use Netty instead of MINA, the strange CPU peak performance problem is gone, so this is not a JDK problem it seems. Same setup (number of IO processor threads, executor threads, number of clients etc etc...) for both Netty and MINA.
I will test again with MINA later to see if this problem is gone by then.
> Strange CPU peak occuring at fixed interval when several thousand connections active
> ------------------------------------------------------------------------------------
>
> Key: DIRMINA-681
> URL: https://issues.apache.org/jira/browse/DIRMINA-681
> Project: MINA
> Issue Type: Task
> Components: Core
> Affects Versions: 2.0.0-M4, 2.0.0-RC1
> Environment: Windows Vista Ultimate 64-bit (on 64-bit Sun JDK 1.6.0_18). Intel Core 2 Quad Core Q9300 2,5 GHz, 8 GB RAM
> Reporter: Mauritz Lovgren
> Fix For: 2.0.0
>
> Attachments: screenshot-1.jpg, screenshot-2.jpg, screenshot-3.jpg, screenshot-4.jpg
>
>
> Observing strange CPU activity occuring at regular (seemingly fixed) interval with no protocol traffic activity.
> See attached window capture of task manager that shows this with 3000 active connections.
> Is there some kind of cleanup occuring within MINA core at a predefined interval?
> The 3000 connections in the example above connects within 250 seconds. A normal situation would be that these connections are established over a longer period of time, perhaps spreading the CPU peaks shown above as well, flattening the curve.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (DIRMINA-681) Strange CPU peak occuring at fixed
interval when several thousand connections active
Posted by "Mauritz Lovgren (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/DIRMINA-681?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12830939#action_12830939 ]
Mauritz Lovgren commented on DIRMINA-681:
-----------------------------------------
FYI, the following occurs while running the load test (as far as I got anyway before stuff stalls)..:
14:20:36,000 WARN Create a new selector. Selected is 0, delta = 0 [NioProcessor-9] (AbstractPollingIoProcessor.java:1075)
14:29:45,788 WARN Create a new selector. Selected is 0, delta = 0 [NioProcessor-10] (AbstractPollingIoProcessor.java:1075)
14:29:45,788 WARN Create a new selector. Selected is 0, delta = 0 [NioProcessor-6] (AbstractPollingIoProcessor.java:1075)
14:29:45,788 WARN Create a new selector. Selected is 0, delta = 0 [NioProcessor-7] (AbstractPollingIoProcessor.java:1075)
14:29:45,803 WARN Create a new selector. Selected is 0, delta = 0 [NioProcessor-6] (AbstractPollingIoProcessor.java:1075)
Further, clients are able to connect at same rate as with RC1 for a few minutes, until suddenly the server (NIO acceptor / connections) seems to reach some strange state, causing message flow to slow down / stop.
> Strange CPU peak occuring at fixed interval when several thousand connections active
> ------------------------------------------------------------------------------------
>
> Key: DIRMINA-681
> URL: https://issues.apache.org/jira/browse/DIRMINA-681
> Project: MINA
> Issue Type: Task
> Components: Core
> Affects Versions: 2.0.0-M4, 2.0.0-RC1
> Environment: Windows Vista Ultimate 64-bit (on 64-bit Sun JDK 1.6.0_18). Intel Core 2 Quad Core Q9300 2,5 GHz, 8 GB RAM
> Reporter: Mauritz Lovgren
> Fix For: 2.0.0
>
> Attachments: screenshot-1.jpg, screenshot-2.jpg, screenshot-3.jpg, screenshot-4.jpg
>
>
> Observing strange CPU activity occuring at regular (seemingly fixed) interval with no protocol traffic activity.
> See attached window capture of task manager that shows this with 3000 active connections.
> Is there some kind of cleanup occuring within MINA core at a predefined interval?
> The 3000 connections in the example above connects within 250 seconds. A normal situation would be that these connections are established over a longer period of time, perhaps spreading the CPU peaks shown above as well, flattening the curve.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (DIRMINA-681) Strange CPU peak occuring at fixed
interval when several thousand connections active
Posted by "Emmanuel Lecharny (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/DIRMINA-681?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12834218#action_12834218 ]
Emmanuel Lecharny commented on DIRMINA-681:
-------------------------------------------
On Windows, the WindowsSelectorImpl class uses an array containing a maximum of 1024 channels. When this number is reached, another one has to be created. This explain why you get those new selector created.
> Strange CPU peak occuring at fixed interval when several thousand connections active
> ------------------------------------------------------------------------------------
>
> Key: DIRMINA-681
> URL: https://issues.apache.org/jira/browse/DIRMINA-681
> Project: MINA
> Issue Type: Task
> Components: Core
> Affects Versions: 2.0.0-M4, 2.0.0-RC1
> Environment: Windows Vista Ultimate 64-bit (on 64-bit Sun JDK 1.6.0_18). Intel Core 2 Quad Core Q9300 2,5 GHz, 8 GB RAM
> Reporter: Mauritz Lovgren
> Fix For: 2.0.0
>
> Attachments: screenshot-1.jpg, screenshot-2.jpg, screenshot-3.jpg, screenshot-4.jpg
>
>
> Observing strange CPU activity occuring at regular (seemingly fixed) interval with no protocol traffic activity.
> See attached window capture of task manager that shows this with 3000 active connections.
> Is there some kind of cleanup occuring within MINA core at a predefined interval?
> The 3000 connections in the example above connects within 250 seconds. A normal situation would be that these connections are established over a longer period of time, perhaps spreading the CPU peaks shown above as well, flattening the curve.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Issue Comment Edited: (DIRMINA-681) Strange CPU peak
occuring at fixed interval when several thousand connections active
Posted by "Mauritz Lovgren (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/DIRMINA-681?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12694479#action_12694479 ]
Mauritz Lovgren edited comment on DIRMINA-681 at 4/1/09 4:32 AM:
-----------------------------------------------------------------
This is easily reproducable by creating a test server having thousands of clients doing almost nothing (just sending a heartbeat or similar every minute). The CPU peak seems to occur every 10 - 12 minutes, lasting about 2 minutes each time it occurs. Highest CPU at the beginning, creating a classic Mount Everest profile :-).
was (Author: mauritz):
This is easily reproducable by creating a test server having thousands of clients doing almost nothing (just sending a heartbeat or similar every minute). The CPU peak seems to occur every 10 - 12 minutes.
> Strange CPU peak occuring at fixed interval when several thousand connections active
> ------------------------------------------------------------------------------------
>
> Key: DIRMINA-681
> URL: https://issues.apache.org/jira/browse/DIRMINA-681
> Project: MINA
> Issue Type: Bug
> Components: Core
> Affects Versions: 2.0.0-M4
> Environment: Windows Vista Ultimate 64-bit (on 64-bit Sun JDK 1.6.0_12). Intel Core 2 Quad Core Q9300 2,5 GHz, 8 GB RAM
> Reporter: Mauritz Lovgren
> Attachments: screenshot-1.jpg, screenshot-2.jpg, screenshot-3.jpg
>
>
> Observing strange CPU activity occuring at regular (seemingly fixed) interval with no protocol traffic activity.
> See attached window capture of task manager that shows this with 3000 active connections.
> Is there some kind of cleanup occuring within MINA core at a predefined interval?
> The 3000 connections in the example above connects within 250 seconds. A normal situation would be that these connections are established over a longer period of time, perhaps spreading the CPU peaks shown above as well, flattening the curve.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (DIRMINA-681) Strange CPU peak occuring at fixed
interval when several thousand connections active
Posted by "Mauritz Lovgren (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/DIRMINA-681?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Mauritz Lovgren updated DIRMINA-681:
------------------------------------
Environment: Windows Vista Ultimate 64-bit (on 64-bit Sun JDK 1.6.0_12). Intel Core 2 Quad Core Q9300 2,5 GHz, 8 GB RAM (was: Windows Vista Ultimate 64-bit (on 64-bit Sun JDK 1.6.0_12))
> Strange CPU peak occuring at fixed interval when several thousand connections active
> ------------------------------------------------------------------------------------
>
> Key: DIRMINA-681
> URL: https://issues.apache.org/jira/browse/DIRMINA-681
> Project: MINA
> Issue Type: Bug
> Components: Core
> Affects Versions: 2.0.0-M4
> Environment: Windows Vista Ultimate 64-bit (on 64-bit Sun JDK 1.6.0_12). Intel Core 2 Quad Core Q9300 2,5 GHz, 8 GB RAM
> Reporter: Mauritz Lovgren
> Attachments: screenshot-1.jpg, screenshot-2.jpg, screenshot-3.jpg
>
>
> Observing strange CPU activity occuring at regular (seemingly fixed) interval with no protocol traffic activity.
> See attached window capture of task manager that shows this with 3000 active connections.
> Is there some kind of cleanup occuring within MINA core at a predefined interval?
> The 3000 connections in the example above connects within 250 seconds. A normal situation would be that these connections are established over a longer period of time, perhaps spreading the CPU peaks shown above as well, flattening the curve.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (DIRMINA-681) Strange CPU peak occuring at fixed
interval when several thousand connections active
Posted by "Emmanuel Lecharny (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/DIRMINA-681?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Emmanuel Lecharny updated DIRMINA-681:
--------------------------------------
Fix Version/s: (was: 2.0.1)
2.0.2
> Strange CPU peak occuring at fixed interval when several thousand connections active
> ------------------------------------------------------------------------------------
>
> Key: DIRMINA-681
> URL: https://issues.apache.org/jira/browse/DIRMINA-681
> Project: MINA
> Issue Type: Task
> Components: Core
> Affects Versions: 2.0.0-M4, 2.0.0-RC1
> Environment: Windows Vista Ultimate 64-bit (on 64-bit Sun JDK 1.6.0_18). Intel Core 2 Quad Core Q9300 2,5 GHz, 8 GB RAM
> Reporter: Mauritz Lovgren
> Fix For: 2.0.2
>
> Attachments: screenshot-1.jpg, screenshot-2.jpg, screenshot-3.jpg, screenshot-4.jpg
>
>
> Observing strange CPU activity occuring at regular (seemingly fixed) interval with no protocol traffic activity.
> See attached window capture of task manager that shows this with 3000 active connections.
> Is there some kind of cleanup occuring within MINA core at a predefined interval?
> The 3000 connections in the example above connects within 250 seconds. A normal situation would be that these connections are established over a longer period of time, perhaps spreading the CPU peaks shown above as well, flattening the curve.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (DIRMINA-681) Strange CPU peak occuring at fixed
interval when several thousand connections active
Posted by "Emmanuel Lecharny (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/DIRMINA-681?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12804957#action_12804957 ]
Emmanuel Lecharny commented on DIRMINA-681:
-------------------------------------------
It would be interesting to gather those informations when running the test :
- CPU consumption (users and system)
- when hitting a CPU pike, generating a thread dump
- adding some GC logs could be interesting too.
> Strange CPU peak occuring at fixed interval when several thousand connections active
> ------------------------------------------------------------------------------------
>
> Key: DIRMINA-681
> URL: https://issues.apache.org/jira/browse/DIRMINA-681
> Project: MINA
> Issue Type: Task
> Components: Core
> Affects Versions: 2.0.0-M4, 2.0.0-RC1
> Environment: Windows Vista Ultimate 64-bit (on 64-bit Sun JDK 1.6.0_18). Intel Core 2 Quad Core Q9300 2,5 GHz, 8 GB RAM
> Reporter: Mauritz Lovgren
> Fix For: 2.0.0
>
> Attachments: screenshot-1.jpg, screenshot-2.jpg, screenshot-3.jpg, screenshot-4.jpg
>
>
> Observing strange CPU activity occuring at regular (seemingly fixed) interval with no protocol traffic activity.
> See attached window capture of task manager that shows this with 3000 active connections.
> Is there some kind of cleanup occuring within MINA core at a predefined interval?
> The 3000 connections in the example above connects within 250 seconds. A normal situation would be that these connections are established over a longer period of time, perhaps spreading the CPU peaks shown above as well, flattening the curve.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (DIRMINA-681) Strange CPU peak occuring at fixed
interval when several thousand connections active
Posted by "Mauritz Lovgren (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/DIRMINA-681?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Mauritz Lovgren updated DIRMINA-681:
------------------------------------
Environment: Windows Vista Ultimate 64-bit (on 64-bit Sun JDK 1.6.0_18). Intel Core 2 Quad Core Q9300 2,5 GHz, 8 GB RAM (was: Windows Vista Ultimate 64-bit (on 64-bit Sun JDK 1.6.0_12). Intel Core 2 Quad Core Q9300 2,5 GHz, 8 GB RAM)
> Strange CPU peak occuring at fixed interval when several thousand connections active
> ------------------------------------------------------------------------------------
>
> Key: DIRMINA-681
> URL: https://issues.apache.org/jira/browse/DIRMINA-681
> Project: MINA
> Issue Type: Task
> Components: Core
> Affects Versions: 2.0.0-M4
> Environment: Windows Vista Ultimate 64-bit (on 64-bit Sun JDK 1.6.0_18). Intel Core 2 Quad Core Q9300 2,5 GHz, 8 GB RAM
> Reporter: Mauritz Lovgren
> Fix For: 2.0.0
>
> Attachments: screenshot-1.jpg, screenshot-2.jpg, screenshot-3.jpg, screenshot-4.jpg
>
>
> Observing strange CPU activity occuring at regular (seemingly fixed) interval with no protocol traffic activity.
> See attached window capture of task manager that shows this with 3000 active connections.
> Is there some kind of cleanup occuring within MINA core at a predefined interval?
> The 3000 connections in the example above connects within 250 seconds. A normal situation would be that these connections are established over a longer period of time, perhaps spreading the CPU peaks shown above as well, flattening the curve.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (DIRMINA-681) Strange CPU peak occuring at fixed
interval when several thousand connections active
Posted by "Mauritz Lovgren (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/DIRMINA-681?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12834210#action_12834210 ]
Mauritz Lovgren commented on DIRMINA-681:
-----------------------------------------
This is fun. I notice that a new WindowsSelectorImpl$SelectorThread is created and started for exactly every 1000 connections. This seems to be a design choice made by Sun for scaling selector for handling groups of thousands of connections. I guess you guys were aware of this already? ;-)
> Strange CPU peak occuring at fixed interval when several thousand connections active
> ------------------------------------------------------------------------------------
>
> Key: DIRMINA-681
> URL: https://issues.apache.org/jira/browse/DIRMINA-681
> Project: MINA
> Issue Type: Task
> Components: Core
> Affects Versions: 2.0.0-M4, 2.0.0-RC1
> Environment: Windows Vista Ultimate 64-bit (on 64-bit Sun JDK 1.6.0_18). Intel Core 2 Quad Core Q9300 2,5 GHz, 8 GB RAM
> Reporter: Mauritz Lovgren
> Fix For: 2.0.0
>
> Attachments: screenshot-1.jpg, screenshot-2.jpg, screenshot-3.jpg, screenshot-4.jpg
>
>
> Observing strange CPU activity occuring at regular (seemingly fixed) interval with no protocol traffic activity.
> See attached window capture of task manager that shows this with 3000 active connections.
> Is there some kind of cleanup occuring within MINA core at a predefined interval?
> The 3000 connections in the example above connects within 250 seconds. A normal situation would be that these connections are established over a longer period of time, perhaps spreading the CPU peaks shown above as well, flattening the curve.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (DIRMINA-681) Strange CPU peak occuring at fixed
interval when several thousand connections active
Posted by "Mauritz Lovgren (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/DIRMINA-681?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12805009#action_12805009 ]
Mauritz Lovgren commented on DIRMINA-681:
-----------------------------------------
I also noticed that a FilterWriteTest (or similar) was hanging during the unit tests, therefore I had to build bypassing unit tests (-Dmaven.test.skip=true). Might be related to the hang problem I notices with RC2? No deadlocks detected though...
> Strange CPU peak occuring at fixed interval when several thousand connections active
> ------------------------------------------------------------------------------------
>
> Key: DIRMINA-681
> URL: https://issues.apache.org/jira/browse/DIRMINA-681
> Project: MINA
> Issue Type: Task
> Components: Core
> Affects Versions: 2.0.0-M4, 2.0.0-RC1
> Environment: Windows Vista Ultimate 64-bit (on 64-bit Sun JDK 1.6.0_18). Intel Core 2 Quad Core Q9300 2,5 GHz, 8 GB RAM
> Reporter: Mauritz Lovgren
> Fix For: 2.0.0
>
> Attachments: screenshot-1.jpg, screenshot-2.jpg, screenshot-3.jpg, screenshot-4.jpg
>
>
> Observing strange CPU activity occuring at regular (seemingly fixed) interval with no protocol traffic activity.
> See attached window capture of task manager that shows this with 3000 active connections.
> Is there some kind of cleanup occuring within MINA core at a predefined interval?
> The 3000 connections in the example above connects within 250 seconds. A normal situation would be that these connections are established over a longer period of time, perhaps spreading the CPU peaks shown above as well, flattening the curve.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (DIRMINA-681) Strange CPU peak occuring at fixed
interval when several thousand connections active
Posted by "Mauritz Lovgren (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/DIRMINA-681?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Mauritz Lovgren updated DIRMINA-681:
------------------------------------
Affects Version/s: 2.0.0-RC1
I conducted profiling on 2.0.0-RC1. (Updated affects versions)
> Strange CPU peak occuring at fixed interval when several thousand connections active
> ------------------------------------------------------------------------------------
>
> Key: DIRMINA-681
> URL: https://issues.apache.org/jira/browse/DIRMINA-681
> Project: MINA
> Issue Type: Task
> Components: Core
> Affects Versions: 2.0.0-M4, 2.0.0-RC1
> Environment: Windows Vista Ultimate 64-bit (on 64-bit Sun JDK 1.6.0_18). Intel Core 2 Quad Core Q9300 2,5 GHz, 8 GB RAM
> Reporter: Mauritz Lovgren
> Fix For: 2.0.0
>
> Attachments: screenshot-1.jpg, screenshot-2.jpg, screenshot-3.jpg, screenshot-4.jpg
>
>
> Observing strange CPU activity occuring at regular (seemingly fixed) interval with no protocol traffic activity.
> See attached window capture of task manager that shows this with 3000 active connections.
> Is there some kind of cleanup occuring within MINA core at a predefined interval?
> The 3000 connections in the example above connects within 250 seconds. A normal situation would be that these connections are established over a longer period of time, perhaps spreading the CPU peaks shown above as well, flattening the curve.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (DIRMINA-681) Strange CPU peak occuring at fixed
interval when several thousand connections active
Posted by "Mauritz Lovgren (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/DIRMINA-681?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12830763#action_12830763 ]
Mauritz Lovgren commented on DIRMINA-681:
-----------------------------------------
I will :-).
> Strange CPU peak occuring at fixed interval when several thousand connections active
> ------------------------------------------------------------------------------------
>
> Key: DIRMINA-681
> URL: https://issues.apache.org/jira/browse/DIRMINA-681
> Project: MINA
> Issue Type: Task
> Components: Core
> Affects Versions: 2.0.0-M4, 2.0.0-RC1
> Environment: Windows Vista Ultimate 64-bit (on 64-bit Sun JDK 1.6.0_18). Intel Core 2 Quad Core Q9300 2,5 GHz, 8 GB RAM
> Reporter: Mauritz Lovgren
> Fix For: 2.0.0
>
> Attachments: screenshot-1.jpg, screenshot-2.jpg, screenshot-3.jpg, screenshot-4.jpg
>
>
> Observing strange CPU activity occuring at regular (seemingly fixed) interval with no protocol traffic activity.
> See attached window capture of task manager that shows this with 3000 active connections.
> Is there some kind of cleanup occuring within MINA core at a predefined interval?
> The 3000 connections in the example above connects within 250 seconds. A normal situation would be that these connections are established over a longer period of time, perhaps spreading the CPU peaks shown above as well, flattening the curve.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (DIRMINA-681) Strange CPU peak occuring at fixed
interval when several thousand connections active
Posted by "Emmanuel Lecharny (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/DIRMINA-681?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12830941#action_12830941 ]
Emmanuel Lecharny commented on DIRMINA-681:
-------------------------------------------
Yes, we have reached the very same state even when the selector is not switched.
Strange enough, after a certain number of exchanged messages, it stall completely. Som trace I got shown that he clients socket are doing nothing. I didn't had tie last week to check what was going on, but this is really strange.
The fix was introduced to remove the blocked selector, and it works, but there is something wrong in another area. There is another JIRA opened for these probelms :
https://issues.apache.org/jira/browse/DIRMINA-762
> Strange CPU peak occuring at fixed interval when several thousand connections active
> ------------------------------------------------------------------------------------
>
> Key: DIRMINA-681
> URL: https://issues.apache.org/jira/browse/DIRMINA-681
> Project: MINA
> Issue Type: Task
> Components: Core
> Affects Versions: 2.0.0-M4, 2.0.0-RC1
> Environment: Windows Vista Ultimate 64-bit (on 64-bit Sun JDK 1.6.0_18). Intel Core 2 Quad Core Q9300 2,5 GHz, 8 GB RAM
> Reporter: Mauritz Lovgren
> Fix For: 2.0.0
>
> Attachments: screenshot-1.jpg, screenshot-2.jpg, screenshot-3.jpg, screenshot-4.jpg
>
>
> Observing strange CPU activity occuring at regular (seemingly fixed) interval with no protocol traffic activity.
> See attached window capture of task manager that shows this with 3000 active connections.
> Is there some kind of cleanup occuring within MINA core at a predefined interval?
> The 3000 connections in the example above connects within 250 seconds. A normal situation would be that these connections are established over a longer period of time, perhaps spreading the CPU peaks shown above as well, flattening the curve.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (DIRMINA-681) Strange CPU peak occuring at fixed
interval when several thousand connections active
Posted by "Mauritz Lovgren (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/DIRMINA-681?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Mauritz Lovgren updated DIRMINA-681:
------------------------------------
Attachment: screenshot-4.jpg
Found that the CPU peak most likely is caused by the code identified by the JProfiler hotspot detector (shown in screenshot-4). I ran the profiling session on JDK 1.6.0_18, which claims to have epoll bug fix backported from JDK 1.7... I assume that this problem might go away with the fixes for MINA 2.0.0-RC2?
> Strange CPU peak occuring at fixed interval when several thousand connections active
> ------------------------------------------------------------------------------------
>
> Key: DIRMINA-681
> URL: https://issues.apache.org/jira/browse/DIRMINA-681
> Project: MINA
> Issue Type: Task
> Components: Core
> Affects Versions: 2.0.0-M4
> Environment: Windows Vista Ultimate 64-bit (on 64-bit Sun JDK 1.6.0_12). Intel Core 2 Quad Core Q9300 2,5 GHz, 8 GB RAM
> Reporter: Mauritz Lovgren
> Fix For: 2.0.0
>
> Attachments: screenshot-1.jpg, screenshot-2.jpg, screenshot-3.jpg, screenshot-4.jpg
>
>
> Observing strange CPU activity occuring at regular (seemingly fixed) interval with no protocol traffic activity.
> See attached window capture of task manager that shows this with 3000 active connections.
> Is there some kind of cleanup occuring within MINA core at a predefined interval?
> The 3000 connections in the example above connects within 250 seconds. A normal situation would be that these connections are established over a longer period of time, perhaps spreading the CPU peaks shown above as well, flattening the curve.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (DIRMINA-681) Strange CPU peak occuring at fixed
interval when several thousand connections active
Posted by "Emmanuel Lecharny (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/DIRMINA-681?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12833789#action_12833789 ]
Emmanuel Lecharny commented on DIRMINA-681:
-------------------------------------------
No. I simply fixed the patch :)
I have mad a mistake : to detect if the selector was dead, I checked 3 things :
- first, the select(1000) should return 0
- second, it should return 0 immediately
- third, it should not have been woke up
For the third condition, I used a boolean flag which was set to true when the IoProcessor was waken up, but in this case, I forgot to reset it after the test (ie, when select retur 0 in 0 ms when it has been awakened). The consequence was that we just return back to the select(1000), doing nothing at all, forever.
I fixed that .
> Strange CPU peak occuring at fixed interval when several thousand connections active
> ------------------------------------------------------------------------------------
>
> Key: DIRMINA-681
> URL: https://issues.apache.org/jira/browse/DIRMINA-681
> Project: MINA
> Issue Type: Task
> Components: Core
> Affects Versions: 2.0.0-M4, 2.0.0-RC1
> Environment: Windows Vista Ultimate 64-bit (on 64-bit Sun JDK 1.6.0_18). Intel Core 2 Quad Core Q9300 2,5 GHz, 8 GB RAM
> Reporter: Mauritz Lovgren
> Fix For: 2.0.0
>
> Attachments: screenshot-1.jpg, screenshot-2.jpg, screenshot-3.jpg, screenshot-4.jpg
>
>
> Observing strange CPU activity occuring at regular (seemingly fixed) interval with no protocol traffic activity.
> See attached window capture of task manager that shows this with 3000 active connections.
> Is there some kind of cleanup occuring within MINA core at a predefined interval?
> The 3000 connections in the example above connects within 250 seconds. A normal situation would be that these connections are established over a longer period of time, perhaps spreading the CPU peaks shown above as well, flattening the curve.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (DIRMINA-681) Strange CPU peak occuring at fixed
interval when several thousand connections active
Posted by "Emmanuel Lecharny (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/DIRMINA-681?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12805005#action_12805005 ]
Emmanuel Lecharny commented on DIRMINA-681:
-------------------------------------------
ok. I have modified the way the select() method works : if at some point we get a spinning epoll, then the new code is supposed to delete the selector() and to create a new one, then reattach all the keys to this new selector.
There might be some bug in this area. I will have a look and see what can go wrong.
> Strange CPU peak occuring at fixed interval when several thousand connections active
> ------------------------------------------------------------------------------------
>
> Key: DIRMINA-681
> URL: https://issues.apache.org/jira/browse/DIRMINA-681
> Project: MINA
> Issue Type: Task
> Components: Core
> Affects Versions: 2.0.0-M4, 2.0.0-RC1
> Environment: Windows Vista Ultimate 64-bit (on 64-bit Sun JDK 1.6.0_18). Intel Core 2 Quad Core Q9300 2,5 GHz, 8 GB RAM
> Reporter: Mauritz Lovgren
> Fix For: 2.0.0
>
> Attachments: screenshot-1.jpg, screenshot-2.jpg, screenshot-3.jpg, screenshot-4.jpg
>
>
> Observing strange CPU activity occuring at regular (seemingly fixed) interval with no protocol traffic activity.
> See attached window capture of task manager that shows this with 3000 active connections.
> Is there some kind of cleanup occuring within MINA core at a predefined interval?
> The 3000 connections in the example above connects within 250 seconds. A normal situation would be that these connections are established over a longer period of time, perhaps spreading the CPU peaks shown above as well, flattening the curve.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (DIRMINA-681) Strange CPU peak occuring at fixed
interval when several thousand connections active
Posted by "Mauritz Lovgren (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/DIRMINA-681?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12834193#action_12834193 ]
Mauritz Lovgren commented on DIRMINA-681:
-----------------------------------------
Tested on latest build of JDK 1.7.0 (ea b83) as well. Same behaviour as with JDK 1.6.0_18.
> Strange CPU peak occuring at fixed interval when several thousand connections active
> ------------------------------------------------------------------------------------
>
> Key: DIRMINA-681
> URL: https://issues.apache.org/jira/browse/DIRMINA-681
> Project: MINA
> Issue Type: Task
> Components: Core
> Affects Versions: 2.0.0-M4, 2.0.0-RC1
> Environment: Windows Vista Ultimate 64-bit (on 64-bit Sun JDK 1.6.0_18). Intel Core 2 Quad Core Q9300 2,5 GHz, 8 GB RAM
> Reporter: Mauritz Lovgren
> Fix For: 2.0.0
>
> Attachments: screenshot-1.jpg, screenshot-2.jpg, screenshot-3.jpg, screenshot-4.jpg
>
>
> Observing strange CPU activity occuring at regular (seemingly fixed) interval with no protocol traffic activity.
> See attached window capture of task manager that shows this with 3000 active connections.
> Is there some kind of cleanup occuring within MINA core at a predefined interval?
> The 3000 connections in the example above connects within 250 seconds. A normal situation would be that these connections are established over a longer period of time, perhaps spreading the CPU peaks shown above as well, flattening the curve.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (DIRMINA-681) Strange CPU peak occuring at fixed
interval when several thousand connections active
Posted by "Mauritz Lovgren (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/DIRMINA-681?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12830938#action_12830938 ]
Mauritz Lovgren commented on DIRMINA-681:
-----------------------------------------
Hm. Still having problems running on the RC2-SNAPSHOT (from trunk as of today). Running on RC1 (final), no problems (except the strange CPU peak as described in the issue description every 10-15 minutes).
Running on RC2-SNAPSHOT, messages time out during delivery between server and client. They seem to stall, whereafter no traffic passes through the proxy at all. Switching back to RC1 and the problem is gone.
When running the test to prove that the CPU peak issue, I need at least a couple of thousand client connections. Currently, the proxy stops handling messages after a few hundred (not been able to connect more than 1000 clients through the proxy with RC2 yet before everything stalls).
Could there be a performance issue with your fix?
> Strange CPU peak occuring at fixed interval when several thousand connections active
> ------------------------------------------------------------------------------------
>
> Key: DIRMINA-681
> URL: https://issues.apache.org/jira/browse/DIRMINA-681
> Project: MINA
> Issue Type: Task
> Components: Core
> Affects Versions: 2.0.0-M4, 2.0.0-RC1
> Environment: Windows Vista Ultimate 64-bit (on 64-bit Sun JDK 1.6.0_18). Intel Core 2 Quad Core Q9300 2,5 GHz, 8 GB RAM
> Reporter: Mauritz Lovgren
> Fix For: 2.0.0
>
> Attachments: screenshot-1.jpg, screenshot-2.jpg, screenshot-3.jpg, screenshot-4.jpg
>
>
> Observing strange CPU activity occuring at regular (seemingly fixed) interval with no protocol traffic activity.
> See attached window capture of task manager that shows this with 3000 active connections.
> Is there some kind of cleanup occuring within MINA core at a predefined interval?
> The 3000 connections in the example above connects within 250 seconds. A normal situation would be that these connections are established over a longer period of time, perhaps spreading the CPU peaks shown above as well, flattening the curve.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (DIRMINA-681) Strange CPU peak occuring at fixed
interval when several thousand connections active
Posted by "Mauritz Lovgren (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/DIRMINA-681?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12830948#action_12830948 ]
Mauritz Lovgren commented on DIRMINA-681:
-----------------------------------------
Yes, seems to me as well that it might perhaps not be the epoll situation that causes this stall.
FYI: There is little CPU ativity when the stall occurs, so it doesn't seem to be a loop issue, more a state problem.
> Strange CPU peak occuring at fixed interval when several thousand connections active
> ------------------------------------------------------------------------------------
>
> Key: DIRMINA-681
> URL: https://issues.apache.org/jira/browse/DIRMINA-681
> Project: MINA
> Issue Type: Task
> Components: Core
> Affects Versions: 2.0.0-M4, 2.0.0-RC1
> Environment: Windows Vista Ultimate 64-bit (on 64-bit Sun JDK 1.6.0_18). Intel Core 2 Quad Core Q9300 2,5 GHz, 8 GB RAM
> Reporter: Mauritz Lovgren
> Fix For: 2.0.0
>
> Attachments: screenshot-1.jpg, screenshot-2.jpg, screenshot-3.jpg, screenshot-4.jpg
>
>
> Observing strange CPU activity occuring at regular (seemingly fixed) interval with no protocol traffic activity.
> See attached window capture of task manager that shows this with 3000 active connections.
> Is there some kind of cleanup occuring within MINA core at a predefined interval?
> The 3000 connections in the example above connects within 250 seconds. A normal situation would be that these connections are established over a longer period of time, perhaps spreading the CPU peaks shown above as well, flattening the curve.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (DIRMINA-681) Strange CPU peak occuring at fixed
interval when several thousand connections active
Posted by "Emmanuel Lecharny (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/DIRMINA-681?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Emmanuel Lecharny updated DIRMINA-681:
--------------------------------------
Fix Version/s: 2.0.1
(was: 2.0.0)
Postponed to 2.0.1
> Strange CPU peak occuring at fixed interval when several thousand connections active
> ------------------------------------------------------------------------------------
>
> Key: DIRMINA-681
> URL: https://issues.apache.org/jira/browse/DIRMINA-681
> Project: MINA
> Issue Type: Task
> Components: Core
> Affects Versions: 2.0.0-M4, 2.0.0-RC1
> Environment: Windows Vista Ultimate 64-bit (on 64-bit Sun JDK 1.6.0_18). Intel Core 2 Quad Core Q9300 2,5 GHz, 8 GB RAM
> Reporter: Mauritz Lovgren
> Fix For: 2.0.1
>
> Attachments: screenshot-1.jpg, screenshot-2.jpg, screenshot-3.jpg, screenshot-4.jpg
>
>
> Observing strange CPU activity occuring at regular (seemingly fixed) interval with no protocol traffic activity.
> See attached window capture of task manager that shows this with 3000 active connections.
> Is there some kind of cleanup occuring within MINA core at a predefined interval?
> The 3000 connections in the example above connects within 250 seconds. A normal situation would be that these connections are established over a longer period of time, perhaps spreading the CPU peaks shown above as well, flattening the curve.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.