[jira] Created: (TAPESTRY-2530) Possible Performance Bottleneck in PageSource.makeObject()

[Tapestry - Dev mailing list]

Possible Performance Bottleneck in PageSource.makeObject()
----------------------------------------------------------

                 Key: TAPESTRY-2530
                 URL: https://issues.apache.org/jira/browse/TAPESTRY-2530
             Project: Tapestry
          Issue Type: Bug
    Affects Versions: 4.1.2
         Environment: JDK 1.5.0_14 64 bit, Tomcat 5.5.26
            Reporter: Stefan Koelmann
            Priority: Critical


We recently migrated from Tapestry 3 to Tapestry 4.1.2 and are experiencing severe performance problems with mediocre load.

At some point our servers start creating new threads until the Tomcat reaches the maximum of 500 live threads. When examining the thread dump, we find one thread holding a lock at the GenericKeyedObjectPool, while hundreds of other threads are waiting to aquire just this lock.

It seems the locking thread is constructig a page from a synchronized context and needs some time to do so, and this leads to the complete application locking up in no time. See the threaddump attached.

We already tried an upgrade to commons-pool 1.4 with no real effect.

--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.


---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@...
For additional commands, e-mail: dev-help@...

[Tapestry - Dev mailing list]

     [ https://issues.apache.org/jira/browse/TAPESTRY-2530?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Stefan Koelmann updated TAPESTRY-2530:
--------------------------------------

    Attachment: threaddump.zip

Thread dump illustrating the problem with the locked threads.

> Possible Performance Bottleneck in PageSource.makeObject()
> ----------------------------------------------------------
>
>                 Key: TAPESTRY-2530
>                 URL: https://issues.apache.org/jira/browse/TAPESTRY-2530
>             Project: Tapestry
>          Issue Type: Bug
>    Affects Versions: 4.1.2
>         Environment: JDK 1.5.0_14 64 bit, Tomcat 5.5.26
>            Reporter: Stefan Koelmann
>            Priority: Critical
>         Attachments: threaddump.zip
>
>
> We recently migrated from Tapestry 3 to Tapestry 4.1.2 and are experiencing severe performance problems with mediocre load.
> At some point our servers start creating new threads until the Tomcat reaches the maximum of 500 live threads. When examining the thread dump, we find one thread holding a lock at the GenericKeyedObjectPool, while hundreds of other threads are waiting to aquire just this lock.
> It seems the locking thread is constructig a page from a synchronized context and needs some time to do so, and this leads to the complete application locking up in no time. See the threaddump attached.
> We already tried an upgrade to commons-pool 1.4 with no real effect.

--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.


---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@...
For additional commands, e-mail: dev-help@...

[Tapestry - Dev mailing list]

     [ https://issues.apache.org/jira/browse/TAPESTRY-2530?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Stefan Koelmann updated TAPESTRY-2530:
--------------------------------------

    Summary: Possible performance bottleneck in PageSource.makeObject()  (was: Possible Performance Bottleneck in PageSource.makeObject())

> Possible performance bottleneck in PageSource.makeObject()
> ----------------------------------------------------------
>
>                 Key: TAPESTRY-2530
>                 URL: https://issues.apache.org/jira/browse/TAPESTRY-2530
>             Project: Tapestry
>          Issue Type: Bug
>    Affects Versions: 4.1.2
>         Environment: JDK 1.5.0_14 64 bit, Tomcat 5.5.26
>            Reporter: Stefan Koelmann
>            Priority: Critical
>         Attachments: threaddump.zip
>
>
> We recently migrated from Tapestry 3 to Tapestry 4.1.2 and are experiencing severe performance problems with mediocre load.
> At some point our servers start creating new threads until the Tomcat reaches the maximum of 500 live threads. When examining the thread dump, we find one thread holding a lock at the GenericKeyedObjectPool, while hundreds of other threads are waiting to aquire just this lock.
> It seems the locking thread is constructig a page from a synchronized context and needs some time to do so, and this leads to the complete application locking up in no time. See the threaddump attached.
> We already tried an upgrade to commons-pool 1.4 with no real effect.

--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.


---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@...
For additional commands, e-mail: dev-help@...

[Tapestry - Dev mailing list]

    [ https://issues.apache.org/jira/browse/TAPESTRY-2530?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12615653#action_12615653 ]

Andreas Andreou commented on TAPESTRY-2530:
-------------------------------------------

Do you know / is it possible to check with T4.1.5 or T4.1.6-SNAPSHOT (which also update the versions of several dependencies)?

I think i remember some performance problems mentioned in the ML (in certain situations) a year ago with 4.1.2,
but 4.1.3 had fixed on them.

> Possible performance bottleneck in PageSource.makeObject()
> ----------------------------------------------------------
>
>                 Key: TAPESTRY-2530
>                 URL: https://issues.apache.org/jira/browse/TAPESTRY-2530
>             Project: Tapestry
>          Issue Type: Bug
>    Affects Versions: 4.1.2
>         Environment: JDK 1.5.0_14 64 bit, Tomcat 5.5.26
>            Reporter: Stefan Koelmann
>            Priority: Critical
>         Attachments: threaddump.zip
>
>
> We recently migrated from Tapestry 3 to Tapestry 4.1.2 and are experiencing severe performance problems with mediocre load.
> At some point our servers start creating new threads until the Tomcat reaches the maximum of 500 live threads. When examining the thread dump, we find one thread holding a lock at the GenericKeyedObjectPool, while hundreds of other threads are waiting to aquire just this lock.
> It seems the locking thread is constructig a page from a synchronized context and needs some time to do so, and this leads to the complete application locking up in no time. See the threaddump attached.
> We already tried an upgrade to commons-pool 1.4 with no real effect.

--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.


---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@...
For additional commands, e-mail: dev-help@...

[Tapestry - Dev mailing list]

    [ https://issues.apache.org/jira/browse/TAPESTRY-2530?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12615988#action_12615988 ]

Stefan Koelmann commented on TAPESTRY-2530:
-------------------------------------------

Upgrading the application to 4.1.5 would be some effort. We once tried, but were getting OGNL-Exceptions, and then had to stick with the old version because time was pressing. Since we were not able to reproduce the issue in a testing environment yet, we would only be able to check the new version in production.

I did not find any performance-related bugs in Release 4.1.3 or 4.1.5. The only interesting bug I found in the mailing list was Hivemind-162, which was fixed in Hivemind 1.1.1 - already used by T4.1.2

> Possible performance bottleneck in PageSource.makeObject()
> ----------------------------------------------------------
>
>                 Key: TAPESTRY-2530
>                 URL: https://issues.apache.org/jira/browse/TAPESTRY-2530
>             Project: Tapestry
>          Issue Type: Bug
>    Affects Versions: 4.1.2
>         Environment: JDK 1.5.0_14 64 bit, Tomcat 5.5.26
>            Reporter: Stefan Koelmann
>            Priority: Critical
>         Attachments: threaddump.zip
>
>
> We recently migrated from Tapestry 3 to Tapestry 4.1.2 and are experiencing severe performance problems with mediocre load.
> At some point our servers start creating new threads until the Tomcat reaches the maximum of 500 live threads. When examining the thread dump, we find one thread holding a lock at the GenericKeyedObjectPool, while hundreds of other threads are waiting to aquire just this lock.
> It seems the locking thread is constructig a page from a synchronized context and needs some time to do so, and this leads to the complete application locking up in no time. See the threaddump attached.
> We already tried an upgrade to commons-pool 1.4 with no real effect.

--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.


---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@...
For additional commands, e-mail: dev-help@...

[Tapestry - Dev mailing list]

    [ https://issues.apache.org/jira/browse/TAPESTRY-2530?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12616024#action_12616024 ]

Andreas Andreou commented on TAPESTRY-2530:
-------------------------------------------

I went over the commit logs - possible performance related changes between 4.1.3 and now are at revisions:
r595338, r613299, r614722 and r646069

Looking at them from http://svn.apache.org/viewvc?view=rev&revision=646069

> Possible performance bottleneck in PageSource.makeObject()
> ----------------------------------------------------------
>
>                 Key: TAPESTRY-2530
>                 URL: https://issues.apache.org/jira/browse/TAPESTRY-2530
>             Project: Tapestry
>          Issue Type: Bug
>    Affects Versions: 4.1.2
>         Environment: JDK 1.5.0_14 64 bit, Tomcat 5.5.26
>            Reporter: Stefan Koelmann
>            Priority: Critical
>         Attachments: threaddump.zip
>
>
> We recently migrated from Tapestry 3 to Tapestry 4.1.2 and are experiencing severe performance problems with mediocre load.
> At some point our servers start creating new threads until the Tomcat reaches the maximum of 500 live threads. When examining the thread dump, we find one thread holding a lock at the GenericKeyedObjectPool, while hundreds of other threads are waiting to aquire just this lock.
> It seems the locking thread is constructig a page from a synchronized context and needs some time to do so, and this leads to the complete application locking up in no time. See the threaddump attached.
> We already tried an upgrade to commons-pool 1.4 with no real effect.

--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.


---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@...
For additional commands, e-mail: dev-help@...

[Tapestry - Dev mailing list]

     [ https://issues.apache.org/jira/browse/TAPESTRY-2530?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Stefan Koelmann updated TAPESTRY-2530:
--------------------------------------

    Attachment: dump_415.zip

Hello,

we were able to reproduce this problem on our testing machine. Although the machine does not run out of threads, we can see that a lot of threads get blocked at the GenericKeyedObjectPool and get stuck there. Some threads are blocked for more than 90 seconds, which would lead to users refreshing the page and making the situation even worse.

An additional problem seems to be that one has to aquire the lock at the object pool to return a page to the tool; so if one thread renders a page; the whole application gets stuck immediately.

We also retested with Tapestry 4.1.5, it does not solve the problem. I attached a threaddump of the tapestry 4.1.5 version, if you need more information, don't bother to ask.

> Possible performance bottleneck in PageSource.makeObject()
> ----------------------------------------------------------
>
>                 Key: TAPESTRY-2530
>                 URL: https://issues.apache.org/jira/browse/TAPESTRY-2530
>             Project: Tapestry
>          Issue Type: Bug
>    Affects Versions: 4.1.2, 4.1.5
>         Environment: JDK 1.5.0_14 64 bit, Tomcat 5.5.26
>            Reporter: Stefan Koelmann
>            Priority: Critical
>         Attachments: dump_415.zip, threaddump.zip
>
>
> We recently migrated from Tapestry 3 to Tapestry 4.1.2 and are experiencing severe performance problems with mediocre load.
> At some point our servers start creating new threads until the Tomcat reaches the maximum of 500 live threads. When examining the thread dump, we find one thread holding a lock at the GenericKeyedObjectPool, while hundreds of other threads are waiting to aquire just this lock.
> It seems the locking thread is constructig a page from a synchronized context and needs some time to do so, and this leads to the complete application locking up in no time. See the threaddump attached.
> We already tried an upgrade to commons-pool 1.4 with no real effect.

--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.


---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@...
For additional commands, e-mail: dev-help@...

[Tapestry - Dev mailing list]

     [ https://issues.apache.org/jira/browse/TAPESTRY-2530?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Stefan Koelmann updated TAPESTRY-2530:
--------------------------------------

    Affects Version/s: 4.1.5

> Possible performance bottleneck in PageSource.makeObject()
> ----------------------------------------------------------
>
>                 Key: TAPESTRY-2530
>                 URL: https://issues.apache.org/jira/browse/TAPESTRY-2530
>             Project: Tapestry
>          Issue Type: Bug
>    Affects Versions: 4.1.2, 4.1.5
>         Environment: JDK 1.5.0_14 64 bit, Tomcat 5.5.26
>            Reporter: Stefan Koelmann
>            Priority: Critical
>         Attachments: dump_415.zip, threaddump.zip
>
>
> We recently migrated from Tapestry 3 to Tapestry 4.1.2 and are experiencing severe performance problems with mediocre load.
> At some point our servers start creating new threads until the Tomcat reaches the maximum of 500 live threads. When examining the thread dump, we find one thread holding a lock at the GenericKeyedObjectPool, while hundreds of other threads are waiting to aquire just this lock.
> It seems the locking thread is constructig a page from a synchronized context and needs some time to do so, and this leads to the complete application locking up in no time. See the threaddump attached.
> We already tried an upgrade to commons-pool 1.4 with no real effect.

--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.


---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@...
For additional commands, e-mail: dev-help@...

[Tapestry - Dev mailing list]

    [ https://issues.apache.org/jira/browse/TAPESTRY-2530?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12616494#action_12616494 ]

Andreas Andreou commented on TAPESTRY-2530:
-------------------------------------------

Is it possible that you've limited the maximum number of total active pages in the pool?

http://tapestry.apache.org/tapestry4.1/developmentguide/performancetuning.html

> Possible performance bottleneck in PageSource.makeObject()
> ----------------------------------------------------------
>
>                 Key: TAPESTRY-2530
>                 URL: https://issues.apache.org/jira/browse/TAPESTRY-2530
>             Project: Tapestry
>          Issue Type: Bug
>    Affects Versions: 4.1.2, 4.1.5
>         Environment: JDK 1.5.0_14 64 bit, Tomcat 5.5.26
>            Reporter: Stefan Koelmann
>            Priority: Critical
>         Attachments: dump_415.zip, threaddump.zip
>
>
> We recently migrated from Tapestry 3 to Tapestry 4.1.2 and are experiencing severe performance problems with mediocre load.
> At some point our servers start creating new threads until the Tomcat reaches the maximum of 500 live threads. When examining the thread dump, we find one thread holding a lock at the GenericKeyedObjectPool, while hundreds of other threads are waiting to aquire just this lock.
> It seems the locking thread is constructig a page from a synchronized context and needs some time to do so, and this leads to the complete application locking up in no time. See the threaddump attached.
> We already tried an upgrade to commons-pool 1.4 with no real effect.

--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.


---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@...
For additional commands, e-mail: dev-help@...

[Tapestry - Dev mailing list]

    [ https://issues.apache.org/jira/browse/TAPESTRY-2530?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12616498#action_12616498 ]

Stefan Koelmann commented on TAPESTRY-2530:
-------------------------------------------

We are using the default settings. I guess that are those defined in the tapestry.page.xml; if so, the max-active and max-idle values are set to -1 (unlimited).

> Possible performance bottleneck in PageSource.makeObject()
> ----------------------------------------------------------
>
>                 Key: TAPESTRY-2530
>                 URL: https://issues.apache.org/jira/browse/TAPESTRY-2530
>             Project: Tapestry
>          Issue Type: Bug
>    Affects Versions: 4.1.2, 4.1.5
>         Environment: JDK 1.5.0_14 64 bit, Tomcat 5.5.26
>            Reporter: Stefan Koelmann
>            Priority: Critical
>         Attachments: dump_415.zip, threaddump.zip
>
>
> We recently migrated from Tapestry 3 to Tapestry 4.1.2 and are experiencing severe performance problems with mediocre load.
> At some point our servers start creating new threads until the Tomcat reaches the maximum of 500 live threads. When examining the thread dump, we find one thread holding a lock at the GenericKeyedObjectPool, while hundreds of other threads are waiting to aquire just this lock.
> It seems the locking thread is constructig a page from a synchronized context and needs some time to do so, and this leads to the complete application locking up in no time. See the threaddump attached.
> We already tried an upgrade to commons-pool 1.4 with no real effect.

--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.


---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@...
For additional commands, e-mail: dev-help@...

[Tapestry - Dev mailing list]

    [ https://issues.apache.org/jira/browse/TAPESTRY-2530?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12619244#action_12619244 ]

Marcus Schulte commented on TAPESTRY-2530:
------------------------------------------

From looking at dump_415.zip, one can see that thread "http-11080-Processor198" holds a lock on object <0x7bf1ecb0> (the page-pool) while constructing a new instance of that page.
A whole lot of threads can be seen waiting for that exact lock. The reason for this is that all the relevant code in GenericKeyedObjectPool.borrowObject() as well as in GKOP.returnObject() is synchronized on the pool itself.
Thus, the construction of a new page currently blocks any other thread from borrowing *any other* page, existant or not, from the pool. This is obviously unnecessary and I'd expect it to be fatal under very high load.
Stefan, I could provide you with a quick patch without any warranty to try out. Are you interested?

> Possible performance bottleneck in PageSource.makeObject()
> ----------------------------------------------------------
>
>                 Key: TAPESTRY-2530
>                 URL: https://issues.apache.org/jira/browse/TAPESTRY-2530
>             Project: Tapestry
>          Issue Type: Bug
>    Affects Versions: 4.1.2, 4.1.5
>         Environment: JDK 1.5.0_14 64 bit, Tomcat 5.5.26
>            Reporter: Stefan Koelmann
>            Priority: Critical
>         Attachments: dump_415.zip, threaddump.zip
>
>
> We recently migrated from Tapestry 3 to Tapestry 4.1.2 and are experiencing severe performance problems with mediocre load.
> At some point our servers start creating new threads until the Tomcat reaches the maximum of 500 live threads. When examining the thread dump, we find one thread holding a lock at the GenericKeyedObjectPool, while hundreds of other threads are waiting to aquire just this lock.
> It seems the locking thread is constructig a page from a synchronized context and needs some time to do so, and this leads to the complete application locking up in no time. See the threaddump attached.
> We already tried an upgrade to commons-pool 1.4 with no real effect.

--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.


---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@...
For additional commands, e-mail: dev-help@...

[Tapestry - Dev mailing list]

    [ https://issues.apache.org/jira/browse/TAPESTRY-2530?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12619269#action_12619269 ]

Stefan Koelmann commented on TAPESTRY-2530:
-------------------------------------------

Hi Marcus,

you are right, the real bottleneck is located in the GenericKeyedObjectPool. I would be glad to retest with your patch and will post the results asap.

What I do not understand is why other tapestry users are not experiencing the problems we do - there have to be websites built on tapestry with much higher load. Maybe there is a problem with our page-Structure leading to long loading times.

> Possible performance bottleneck in PageSource.makeObject()
> ----------------------------------------------------------
>
>                 Key: TAPESTRY-2530
>                 URL: https://issues.apache.org/jira/browse/TAPESTRY-2530
>             Project: Tapestry
>          Issue Type: Bug
>    Affects Versions: 4.1.2, 4.1.5
>         Environment: JDK 1.5.0_14 64 bit, Tomcat 5.5.26
>            Reporter: Stefan Koelmann
>            Priority: Critical
>         Attachments: dump_415.zip, threaddump.zip
>
>
> We recently migrated from Tapestry 3 to Tapestry 4.1.2 and are experiencing severe performance problems with mediocre load.
> At some point our servers start creating new threads until the Tomcat reaches the maximum of 500 live threads. When examining the thread dump, we find one thread holding a lock at the GenericKeyedObjectPool, while hundreds of other threads are waiting to aquire just this lock.
> It seems the locking thread is constructig a page from a synchronized context and needs some time to do so, and this leads to the complete application locking up in no time. See the threaddump attached.
> We already tried an upgrade to commons-pool 1.4 with no real effect.

--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.


---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@...
For additional commands, e-mail: dev-help@...

[Tapestry - Dev mailing list]

    [ https://issues.apache.org/jira/browse/TAPESTRY-2530?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12619270#action_12619270 ]

Marcus Schulte commented on TAPESTRY-2530:
------------------------------------------

Hi Stefan, yes, maybe you can check your page-loading times by enabling the appropriate logger ... . Nevertheless I'll provide the patch, because there definitively is unnecassary lock-contention which probably impacts all Tap-4 users, if only to a lesser extent.

> Possible performance bottleneck in PageSource.makeObject()
> ----------------------------------------------------------
>
>                 Key: TAPESTRY-2530
>                 URL: https://issues.apache.org/jira/browse/TAPESTRY-2530
>             Project: Tapestry
>          Issue Type: Bug
>    Affects Versions: 4.1.2, 4.1.5
>         Environment: JDK 1.5.0_14 64 bit, Tomcat 5.5.26
>            Reporter: Stefan Koelmann
>            Priority: Critical
>         Attachments: dump_415.zip, threaddump.zip
>
>
> We recently migrated from Tapestry 3 to Tapestry 4.1.2 and are experiencing severe performance problems with mediocre load.
> At some point our servers start creating new threads until the Tomcat reaches the maximum of 500 live threads. When examining the thread dump, we find one thread holding a lock at the GenericKeyedObjectPool, while hundreds of other threads are waiting to aquire just this lock.
> It seems the locking thread is constructig a page from a synchronized context and needs some time to do so, and this leads to the complete application locking up in no time. See the threaddump attached.
> We already tried an upgrade to commons-pool 1.4 with no real effect.

--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.


---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@...
For additional commands, e-mail: dev-help@...

[Tapestry - Dev mailing list]

     [ https://issues.apache.org/jira/browse/TAPESTRY-2530?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Marcus Schulte reassigned TAPESTRY-2530:
----------------------------------------

    Assignee: Marcus Schulte

> Possible performance bottleneck in PageSource.makeObject()
> ----------------------------------------------------------
>
>                 Key: TAPESTRY-2530
>                 URL: https://issues.apache.org/jira/browse/TAPESTRY-2530
>             Project: Tapestry
>          Issue Type: Bug
>    Affects Versions: 4.1.2, 4.1.5
>         Environment: JDK 1.5.0_14 64 bit, Tomcat 5.5.26
>            Reporter: Stefan Koelmann
>            Assignee: Marcus Schulte
>            Priority: Critical
>         Attachments: dump_415.zip, page-pool-patch-1.0-SNAPSHOT.jar, threaddump.zip
>
>
> We recently migrated from Tapestry 3 to Tapestry 4.1.2 and are experiencing severe performance problems with mediocre load.
> At some point our servers start creating new threads until the Tomcat reaches the maximum of 500 live threads. When examining the thread dump, we find one thread holding a lock at the GenericKeyedObjectPool, while hundreds of other threads are waiting to aquire just this lock.
> It seems the locking thread is constructig a page from a synchronized context and needs some time to do so, and this leads to the complete application locking up in no time. See the threaddump attached.
> We already tried an upgrade to commons-pool 1.4 with no real effect.

--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.


---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@...
For additional commands, e-mail: dev-help@...

[Tapestry - Dev mailing list]

     [ https://issues.apache.org/jira/browse/TAPESTRY-2530?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Marcus Schulte updated TAPESTRY-2530:
-------------------------------------

    Attachment: page-pool-patch-1.0-SNAPSHOT.jar

A quick, exploratory fix. Drop into WEB-INF/lib to try it.

> Possible performance bottleneck in PageSource.makeObject()
> ----------------------------------------------------------
>
>                 Key: TAPESTRY-2530
>                 URL: https://issues.apache.org/jira/browse/TAPESTRY-2530
>             Project: Tapestry
>          Issue Type: Bug
>    Affects Versions: 4.1.2, 4.1.5
>         Environment: JDK 1.5.0_14 64 bit, Tomcat 5.5.26
>            Reporter: Stefan Koelmann
>            Priority: Critical
>         Attachments: dump_415.zip, page-pool-patch-1.0-SNAPSHOT.jar, threaddump.zip
>
>
> We recently migrated from Tapestry 3 to Tapestry 4.1.2 and are experiencing severe performance problems with mediocre load.
> At some point our servers start creating new threads until the Tomcat reaches the maximum of 500 live threads. When examining the thread dump, we find one thread holding a lock at the GenericKeyedObjectPool, while hundreds of other threads are waiting to aquire just this lock.
> It seems the locking thread is constructig a page from a synchronized context and needs some time to do so, and this leads to the complete application locking up in no time. See the threaddump attached.
> We already tried an upgrade to commons-pool 1.4 with no real effect.

--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.


---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@...
For additional commands, e-mail: dev-help@...