date:20060626


Hi all,

IMO this is mostly Costin related, because he has something
inside the sandbox :)
Now my question is, could the current sandbox content be
moved to the /sandbox/abc so I (or someone else)
can create a /sandbox/xyz ?

Right now the sandbox is not the 'sandbox' but rather
the Tomcat6Sanbox thought.

Regards,
Mladen.

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

DO NOT REPLY [Bug 38048] - Classloader leak caused by EL evaluation

DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG·
RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT
.
ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND·
INSERTED IN THE BUG DATABASE.

http://issues.apache.org/bugzilla/show_bug.cgi?id=38048





--- Additional Comments From [EMAIL PROTECTED]  2006-06-26 14:18 ---
Created an attachment (id=18537)
 --> (http://issues.apache.org/bugzilla/attachment.cgi?id=18537&action=view)
Modified WebappClassLoader (from 5.5.17) with EL cache cleanup

I've attached a version of WebappClassLoader.java (from 5.5.17) that cleans up
the EL cache (only for classes from the current webapp) on undeploy.


-- 
Configure bugmail: http://issues.apache.org/bugzilla/userprefs.cgi?tab=email
--- You are receiving this mail because: ---
You are the assignee for the bug, or are watching the assignee.

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

DO NOT REPLY [Bug 38048] - Classloader leak caused by EL evaluation

DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG·
RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT
.
ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND·
INSERTED IN THE BUG DATABASE.

http://issues.apache.org/bugzilla/show_bug.cgi?id=38048





--- Additional Comments From [EMAIL PROTECTED]  2006-06-26 14:41 ---
Would it be possible to re-attach as a unified diff format patch ?  Thanks

-- 
Configure bugmail: http://issues.apache.org/bugzilla/userprefs.cgi?tab=email
--- You are receiving this mail because: ---
You are the assignee for the bug, or are watching the assignee.

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

DO NOT REPLY [Bug 38048] - Classloader leak caused by EL evaluation

DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG·
RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT
.
ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND·
INSERTED IN THE BUG DATABASE.

http://issues.apache.org/bugzilla/show_bug.cgi?id=38048





--- Additional Comments From [EMAIL PROTECTED]  2006-06-26 14:45 ---
Created an attachment (id=18538)
 --> (http://issues.apache.org/bugzilla/attachment.cgi?id=18538&action=view)
Patch from 5.5.17


-- 
Configure bugmail: http://issues.apache.org/bugzilla/userprefs.cgi?tab=email
--- You are receiving this mail because: ---
You are the assignee for the bug, or are watching the assignee.

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

DO NOT REPLY [Bug 38048] - Classloader leak caused by EL evaluation

DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG·
RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT
.
ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND·
INSERTED IN THE BUG DATABASE.

http://issues.apache.org/bugzilla/show_bug.cgi?id=38048


[EMAIL PROTECTED] changed:

   What|Removed |Added

Version|5.5.9   |5.5.17




--- Additional Comments From [EMAIL PROTECTED]  2006-06-26 14:49 ---
With the provided patch a number of big webapps with plenty dependencies now can
be properly unloaded, as can be verified using a JVMTI agent (YourKit or other).

With this patch, the only leaks I see are InheritableThreadLocals from Spring.
(attached to lazy-started subsystem threads like the Java2D Disposer)

-- 
Configure bugmail: http://issues.apache.org/bugzilla/userprefs.cgi?tab=email
--- You are receiving this mail because: ---
You are the assignee for the bug, or are watching the assignee.

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Re: HELP Wanted

2006-06-26 Thread Tom Miller

Filip,

I went through the configuration file 
tom-configurations.tar.gz  and could not find any 
configuration refer to multi context (container) 
and multi virtualhost for appache.  I hope, I did not
miss communicated.  Multiple context (container)
using the same Tomcat is one of the main requirement.


Thanks,

Tom


 Original message 
>Date: Tue, 13 Jun 2006 15:05:28 -0400
>From: Tom Miller <[EMAIL PROTECTED]>  
>Subject: HELP Wanted  
>To: users@tomcat.apache.org
>Cc: dev@tomcat.apache.org 
>
>Summarry:
>I have five webased web site with java application.  One 
>application per web site.
>
>Here is my current enviroment:
>OS =  Linux RedHat 4.0
>Appache 2.0.50
>Connector JK 1.2.6
>Tomcat 5.5.4
>
>I am currently have each website and application run on one 
>machine.   I also have some machine with two Tomcat 
>instance.  Each Tomcat instance support one virtual website 
>and one application (context).
>
>
>
>Here is my new requirements:
>
>I have 5 webased application:
>
>
>Main Site:
>I would like all five web site run on the same machine #1 
>(with Apache and JK connector). 
>I would like all five application using the same Tomcat 
>instance (multiple context) on the same machine #2.
>Security and protect for both web site and application 
>container need tobe consider in your implementation.
>
>Failover and Clustering Site:
>I would like machine #3 to be in cluster with machine #1 
and 
>machine #4 to be cluster with machine number #2.
>I will use load balance for the appache (machine #1 and 
>machine #3).
>If you have a better sugguestion, I am listening.
>
>
>I also included a sample topology.
>
>
>
>I have built and configured all of my site with appache, 
>tomcat and connector.  The part I need help with is the 
>multiple context with one Tomcat instance and clustering.
>I am currently have very little time to figure out how 
>multiple context and clustering works.   
>
>This is where I can use some help from some of you who have 
>experienced with it and wanted to make some extra money.   
>Please email if you are interested.
>
>If none of you volunteer I will have to post my need on the 
>list and keep bother you all until I get some solution (-: 
>(-:
>
>[EMAIL PROTECTED]
>
>
>Thanks much,
>
>Tom

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

DO NOT REPLY [Bug 38048] - Classloader leak caused by EL evaluation

DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG·
RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT
.
ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND·
INSERTED IN THE BUG DATABASE.

http://issues.apache.org/bugzilla/show_bug.cgi?id=38048





--- Additional Comments From [EMAIL PROTECTED]  2006-06-26 15:58 ---
Your patch seems ok to me.

-- 
Configure bugmail: http://issues.apache.org/bugzilla/userprefs.cgi?tab=email
--- You are receiving this mail because: ---
You are the assignee for the bug, or are watching the assignee.

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Re: HELP Wanted

2006-06-26 Thread Tom Miller

Please ingore my previous email.  I sent to the wrong people 
and list.  Sorry,

Tom



 Original message 
>Date: Mon, 26 Jun 2006 11:19:22 -0400
>From: Tom Miller <[EMAIL PROTECTED]>  
>Subject: Re: HELP Wanted  
>To: [EMAIL PROTECTED]
>Cc: dev@tomcat.apache.org
>
>Filip,
>
>I went through the configuration file 
>tom-configurations.tar.gz  and could not find any 
>configuration refer to multi context (container) 
>and multi virtualhost for appache.  I hope, I did not
>miss communicated.  Multiple context (container)
>using the same Tomcat is one of the main requirement.
>
>
>Thanks,
>
>Tom
>
>
> Original message 
>>Date: Tue, 13 Jun 2006 15:05:28 -0400
>>From: Tom Miller <[EMAIL PROTECTED]>  
>>Subject: HELP Wanted  
>>To: users@tomcat.apache.org
>>Cc: dev@tomcat.apache.org 
>>
>>Summarry:
>>I have five webased web site with java application.  One 
>>application per web site.
>>
>>Here is my current enviroment:
>>OS =  Linux RedHat 4.0
>>Appache 2.0.50
>>Connector JK 1.2.6
>>Tomcat 5.5.4
>>
>>I am currently have each website and application run on 
one 
>>machine.   I also have some machine with two Tomcat 
>>instance.  Each Tomcat instance support one virtual 
website 
>>and one application (context).
>>
>>
>>
>>Here is my new requirements:
>>
>>I have 5 webased application:
>>
>>
>>Main Site:
>>I would like all five web site run on the same machine #1 
>>(with Apache and JK connector). 
>>I would like all five application using the same Tomcat 
>>instance (multiple context) on the same machine #2.
>>Security and protect for both web site and application 
>>container need tobe consider in your implementation.
>>
>>Failover and Clustering Site:
>>I would like machine #3 to be in cluster with machine #1 
>and 
>>machine #4 to be cluster with machine number #2.
>>I will use load balance for the appache (machine #1 and 
>>machine #3).
>>If you have a better sugguestion, I am listening.
>>
>>
>>I also included a sample topology.
>>
>>
>>
>>I have built and configured all of my site with appache, 
>>tomcat and connector.  The part I need help with is the 
>>multiple context with one Tomcat instance and clustering.
>>I am currently have very little time to figure out how 
>>multiple context and clustering works.   
>>
>>This is where I can use some help from some of you who 
have 
>>experienced with it and wanted to make some extra money.   
>>Please email if you are interested.
>>
>>If none of you volunteer I will have to post my need on 
the 
>>list and keep bother you all until I get some solution (-: 
>>(-:
>>
>>[EMAIL PROTECTED]
>>
>>
>>Thanks much,
>>
>>Tom
>
>
-
>To unsubscribe, e-mail: [EMAIL PROTECTED]
>For additional commands, e-mail: [EMAIL PROTECTED]
>

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Re: Tomcat sandbox

Hi Mladen,

Costins README.txt inside the sandbox already suggested:

===
Sandbox components should be developed in a sub-directory of the /tomcat/sandbox
directory. Developers may choose whether to add source directly under the
component's directory or to create the standard subversion /trunk/, /branches/
and /tags/ directories and add the source under the /trunk/ directory.

The svn commands to do this will look something like:
svn mkdir https://svn.apache.org/repos/asf/tomcat/sandbox/component
svn mkdir https://svn.apache.org/repos/asf/tomcat/sandbox/component/trunk
svn mkdir https://svn.apache.org/repos/asf/tomcat/sandbox/component/branches
svn mkdir https://svn.apache.org/repos/asf/tomcat/sandbox/component/tags

Alternatively, you can use the java/ directory. If possible:
1. create a new package ( or few ) for each module
2. if you have an external dependency for an optional component - create a new
package.
3. Update the top level build.xml file with targets for your module - use
patterns instead
of exclude.

The single tree may help increase visibility of the source code. We are in 2005
( or more ), build tools
and IDEs have very good support for include/exclude patterns, we have tools to
find dependencies between
packages, and we have a single community, so we don't need fragmented source trees and more, at least in
sandbox.

===

+1 for the "components" way of doing it. More precisely to prevent
duplicate trees, component should include a more precise description of
the issue, otherwise we might soon end up blocking general dirs like
"cluster", "connectors" etc.

Regards,

Rainer

Mladen Turk schrieb:

Hi all,

IMO this is mostly Costin related, because he has something
inside the sandbox :)
Now my question is, could the current sandbox content be
moved to the /sandbox/abc so I (or someone else)
can create a /sandbox/xyz ?

Right now the sandbox is not the 'sandbox' but rather
the Tomcat6Sanbox thought.

Regards,
Mladen.

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Re: Tomcat sandbox


Rainer Jung wrote:

Hi Mladen,

Costins README.txt inside the sandbox already suggested:



I know, I read the README, but the content inside does
not follow the README. That's the point :)

Regards,
Mladen.

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Re: NIO vs BIO speed

2006-06-26 Thread devlists

man that sucks,
I can run the same tests, even hundred thousand iterations, I get a little
worse performance than the blocking connector and a little better than the
APR connector, and it never crashes for me.

I'm in dublin at apachecon this week, but will continue as soon I get back,

have a good week!

> Filip Hanik - Dev Lists wrote:
>> yes, I think they are looking pretty good. And I am fairly confident in
>> this new code, as most of it, is old tested APR code.
>
> (Of course, there hasn't been any new changes, so it's not a surprise it
> not working any better for me; I did reboot in the meantime, though, and
> I am running JRE 1.5.0_06)
>
> Performance with NIO is still bad for me (3 times slower than the other
> connectors) with /usr/sbin/ab.exe -n 5000 -c 20 -k
> http://127.0.0.1:8081/tomcat.gif. Throughput seems to vary wildly during
> the test. With /usr/sbin/ab.exe -n 5000 -c 20
> http://127.0.0.1:8081/tomcat.gif (no keepalive), performance is
> horrible, and kills the poller thread after a few hundred iterations
> with the exception I reported earlier:
> Exception in thread "http-8081-Poller-0" java.lang.NullPointerException
>  at sun.nio.ch.WindowsSelectorImpl$FdMap.remove(Unknown Source)
>  at sun.nio.ch.WindowsSelectorImpl$FdMap.access$3000(Unknown
> Source)
>  at sun.nio.ch.WindowsSelectorImpl.implDereg(Unknown Source)
>  at sun.nio.ch.SelectorImpl.processDeregisterQueue(Unknown Source)
>  at sun.nio.ch.WindowsSelectorImpl.doSelect(Unknown Source)
>  at sun.nio.ch.SelectorImpl.lockAndDoSelect(Unknown Source)
>  at sun.nio.ch.SelectorImpl.select(Unknown Source)
>  at
> org.apache.tomcat.util.net.NioEndpoint$Poller.run(NioEndpoint.java:1189)
>  at java.lang.Thread.run(Unknown Source)
>
> Rémy
>
>
> -
> To unsubscribe, e-mail: [EMAIL PROTECTED]
> For additional commands, e-mail: [EMAIL PROTECTED]
>



-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

svn commit: r417257 - /tomcat/tc6.0.x/trunk/java/org/apache/tomcat/util/net/NioEndpoint.java

2006-06-26 Thread remm

Author: remm
Date: Mon Jun 26 10:48:32 2006
New Revision: 417257

URL: http://svn.apache.org/viewvc?rev=417257&view=rev
Log:
- Catch everything, as any unexpected exception would kill the server.

Modified:
tomcat/tc6.0.x/trunk/java/org/apache/tomcat/util/net/NioEndpoint.java

Modified: tomcat/tc6.0.x/trunk/java/org/apache/tomcat/util/net/NioEndpoint.java
URL: 
http://svn.apache.org/viewvc/tomcat/tc6.0.x/trunk/java/org/apache/tomcat/util/net/NioEndpoint.java?rev=417257&r1=417256&r2=417257&view=diff
==
--- tomcat/tc6.0.x/trunk/java/org/apache/tomcat/util/net/NioEndpoint.java 
(original)
+++ tomcat/tc6.0.x/trunk/java/org/apache/tomcat/util/net/NioEndpoint.java Mon 
Jun 26 10:48:32 2006
@@ -1187,7 +1187,7 @@
 int keyCount = 0;
 try {
 keyCount = selector.select(selectorTimeout);
-} catch (IOException x) {
+} catch (Throwable x) {
 log.error("",x);
 continue;
 }



-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

RE: NIO vs BIO speed

2006-06-26 Thread Bill Barker

I haven't tested Filip's connector yet, but that's also my experience with
the AJP/NIO connector:  NIO is pretty much useless on Windows. 

> -Original Message-
> From: Remy Maucherat [mailto:[EMAIL PROTECTED] 
> Sent: Monday, June 26, 2006 4:58 AM
> To: Tomcat Developers List
> Subject: Re: NIO vs BIO speed
> 
> Filip Hanik - Dev Lists wrote:
> > yes, I think they are looking pretty good. And I am fairly 
> confident in 
> > this new code, as most of it, is old tested APR code.
> 
> (Of course, there hasn't been any new changes, so it's not a 
> surprise it 
> not working any better for me; I did reboot in the meantime, 
> though, and 
> I am running JRE 1.5.0_06)
> 
> Performance with NIO is still bad for me (3 times slower than 
> the other 
> connectors) with /usr/sbin/ab.exe -n 5000 -c 20 -k 
> http://127.0.0.1:8081/tomcat.gif. Throughput seems to vary 
> wildly during 
> the test. With /usr/sbin/ab.exe -n 5000 -c 20 
> http://127.0.0.1:8081/tomcat.gif (no keepalive), performance is 
> horrible, and kills the poller thread after a few hundred iterations 
> with the exception I reported earlier:
> Exception in thread "http-8081-Poller-0" 
> java.lang.NullPointerException
>  at 
> sun.nio.ch.WindowsSelectorImpl$FdMap.remove(Unknown Source)
>  at 
> sun.nio.ch.WindowsSelectorImpl$FdMap.access$3000(Unknown Source)
>  at sun.nio.ch.WindowsSelectorImpl.implDereg(Unknown Source)
>  at 
> sun.nio.ch.SelectorImpl.processDeregisterQueue(Unknown Source)
>  at sun.nio.ch.WindowsSelectorImpl.doSelect(Unknown Source)
>  at sun.nio.ch.SelectorImpl.lockAndDoSelect(Unknown Source)
>  at sun.nio.ch.SelectorImpl.select(Unknown Source)
>  at 
> org.apache.tomcat.util.net.NioEndpoint$Poller.run(NioEndpoint.
> java:1189)
>  at java.lang.Thread.run(Unknown Source)
> 
> Rémy
> 
> 
> -
> To unsubscribe, e-mail: [EMAIL PROTECTED]
> For additional commands, e-mail: [EMAIL PROTECTED]
> 
> 
> 

This message is intended only for the use of the person(s) listed above as the 
intended recipient(s), and may contain information that is PRIVILEGED and 
CONFIDENTIAL.  If you are not an intended recipient, you may not read, copy, or 
distribute this message or any attachment. If you received this communication 
in error, please notify us immediately by e-mail and then delete all copies of 
this message and any attachments.

In addition you should be aware that ordinary (unencrypted) e-mail sent through 
the Internet is not secure. Do not send confidential or sensitive information, 
such as social security numbers, account numbers, personal identification 
numbers and passwords, to us via ordinary (unencrypted) e-mail.

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Re: Tomcat sandbox

I should have been more explicit: +1 for cleaning it up according to the 
README :)


Mladen Turk schrieb:

Rainer Jung wrote:

Hi Mladen,

Costins README.txt inside the sandbox already suggested:



I know, I read the README, but the content inside does
not follow the README. That's the point :)


-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

svn commit: r417271 - /tomcat/tc6.0.x/trunk/build.properties.default

2006-06-26 Thread jfclere

Author: jfclere
Date: Mon Jun 26 11:25:51 2006
New Revision: 417271

URL: http://svn.apache.org/viewvc?rev=417271&view=rev
Log:
Change the compiler to the lastest version.

Modified:
tomcat/tc6.0.x/trunk/build.properties.default

Modified: tomcat/tc6.0.x/trunk/build.properties.default
URL: 
http://svn.apache.org/viewvc/tomcat/tc6.0.x/trunk/build.properties.default?rev=417271&r1=417270&r2=417271&view=diff
==
--- tomcat/tc6.0.x/trunk/build.properties.default (original)
+++ tomcat/tc6.0.x/trunk/build.properties.default Mon Jun 26 11:25:51 2006
@@ -34,8 +34,9 @@
 # - Eclipse JDT, version 3.2 or later -
 jdt.home=${base.path}/eclipse/plugins
 jdt.lib=${jdt.home}
-jdt.jar=${jdt.lib}/org.eclipse.jdt.core_3.2.0.v_661.jar
-jdt.loc=http://sunsite.informatik.rwth-aachen.de/eclipse/downloads/drops/S-3.2RC2-200604281315/eclipse-JDT-3.2RC2.zip
+jdt.jar=${jdt.lib}/org.eclipse.jdt.core_3.2.0.v_670.jar
+jdt.loc=http://sunsite.informatik.rwth-aachen.de/eclipse/downloads/drops/S-3.2RC7-200606021317/eclipse-JDT-3.2RC7.zip
+#jdt.loc=http://sunsite.informatik.rwth-aachen.de/eclipse/downloads/drops/S-3.2RC2-200604281315/eclipse-JDT-3.2RC2.zip
 
 # - Tomcat native library -
 tomcat-native.home=${base.path}/tomcat-native-1.1.3



-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Re: svn commit: r416897 - in /tomcat/connectors/trunk/jk/native/common: jk_ajp_common.c jk_ajp_common.h jk_lb_worker.c jk_service.h jk_worker.c


[EMAIL PROTECTED] schrieb:

Author: mturk
Date: Sat Jun 24 03:44:34 2006
New Revision: 416897

URL: http://svn.apache.org/viewvc?rev=416897&view=rev
Log:
Instead calling time(NULL), use it as a function
parameter.
  
Modified: tomcat/connectors/trunk/jk/native/common/jk_lb_worker.c


@@ -283,32 +284,32 @@
 return curmax;
 }
 
-static int JK_METHOD maintain_workers(jk_worker_t *p, jk_logger_t *l)

+static int JK_METHOD maintain_workers(jk_worker_t *p, time_t now, jk_logger_t 
*l)
 {
 unsigned int i = 0;
 jk_uint64_t curmax = 0;
 long delta;
-time_t now = time(NULL);
-JK_TRACE_ENTER(l);
+now = time(NULL);
  

Shouldn't we also drop this last line?


-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Re: svn commit: r416897 - in /tomcat/connectors/trunk/jk/native/common: jk_ajp_common.c jk_ajp_common.h jk_lb_worker.c jk_service.h jk_worker.c


Rainer Jung wrote:
 
-static int JK_METHOD maintain_workers(jk_worker_t *p, jk_logger_t *l)
+static int JK_METHOD maintain_workers(jk_worker_t *p, time_t now, 
jk_logger_t *l)

 {
 unsigned int i = 0;
 jk_uint64_t curmax = 0;
 long delta;
-time_t now = time(NULL);
-JK_TRACE_ENTER(l);
+now = time(NULL);
  

Shouldn't we also drop this last line?



Yes we should. That was the point of having a single
call for time(NULL). Since its resolution is in seconds,
IMHO we are safe with any CPU > 1 MHz :)

Regards,
Mladen.

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

svn commit: r417280 - /tomcat/connectors/trunk/jk/native/common/jk_lb_worker.c

2006-06-26 Thread rjung

Author: rjung
Date: Mon Jun 26 12:10:43 2006
New Revision: 417280

URL: http://svn.apache.org/viewvc?rev=417280&view=rev
Log:
Drop the remaining duplicate stting of now=time(NULL).
Minor Fix to commit r416897.

Modified:
tomcat/connectors/trunk/jk/native/common/jk_lb_worker.c

Modified: tomcat/connectors/trunk/jk/native/common/jk_lb_worker.c
URL: 
http://svn.apache.org/viewvc/tomcat/connectors/trunk/jk/native/common/jk_lb_worker.c?rev=417280&r1=417279&r2=417280&view=diff
==
--- tomcat/connectors/trunk/jk/native/common/jk_lb_worker.c (original)
+++ tomcat/connectors/trunk/jk/native/common/jk_lb_worker.c Mon Jun 26 12:10:43 
2006
@@ -289,7 +289,6 @@
 unsigned int i = 0;
 jk_uint64_t curmax = 0;
 long delta;
-now = time(NULL);
 
 JK_TRACE_ENTER(l);
 if (p && p->worker_private) {



-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Re: NIO vs BIO speed

2006-06-26 Thread Remy Maucherat


Bill Barker wrote:

I haven't tested Filip's connector yet, but that's also my experience with
the AJP/NIO connector:  NIO is pretty much useless on Windows. 


That's what I see, then, but it's not what Sun says.

Rémy

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

jk/native/common/jk_ajp_common.c: Buglet concerning ep_mincache_sz


Hi Mladen,

In jk/native/common/jk_ajp_common.c: Are you able to convince me, that 
this condition is correct:


 2210  if (n > aw->ep_mincache_sz) {
 2211  if (JK_IS_DEBUG_LEVEL(l)) {
 2212  jk_log(l, JK_LOG_DEBUG,
 2213  "reached pool min size %u from %u cache 
slots",

 2214  aw->ep_mincache_sz, aw->ep_cache_sz);
 2215  }
 2216  break;
 2217  }

Variable n counts how many connections we closed during this run of 
maintain and ep_mincache_sz as I understand it, should be the amount of 
connections established (or actually being used).


In order to determine the number of connections we are allowed to close 
we need to know, how many are already closed (or the complement = 
established).


So we must either iterate through the cache two times (first counting 
established or used connections) and then actually closing.
Or we add the number of established connections as a member to the 
worker and only close connections if established > ep_mincache_sz.


Am I wrong?


-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

svn commit: r417288 - /tomcat/connectors/trunk/jk/native/common/jk_ajp_common.c

2006-06-26 Thread mturk

Author: mturk
Date: Mon Jun 26 13:07:54 2006
New Revision: 417288

URL: http://svn.apache.org/viewvc?rev=417288&view=rev
Log:
Properly determine the minimum cache size.

Modified:
tomcat/connectors/trunk/jk/native/common/jk_ajp_common.c

Modified: tomcat/connectors/trunk/jk/native/common/jk_ajp_common.c
URL: 
http://svn.apache.org/viewvc/tomcat/connectors/trunk/jk/native/common/jk_ajp_common.c?rev=417288&r1=417287&r2=417288&view=diff
==
--- tomcat/connectors/trunk/jk/native/common/jk_ajp_common.c (original)
+++ tomcat/connectors/trunk/jk/native/common/jk_ajp_common.c Mon Jun 26 
13:07:54 2006
@@ -2188,7 +2188,12 @@
 }
 JK_ENTER_CS(&aw->cs, rc);
 if (rc) {
-unsigned int i, n = 0;
+unsigned int i, n = 0, cnt = 0;
+/* Count opended slots */
+for (i = 0; i < aw->ep_cache_sz; i++) {
+if (aw->ep_cache[i] && aw->ep_cache[i]->sd >= 0)
+cnt++;
+}
 /* Handle worker cache and recycle timeouts */
 for (i = 0; i < aw->ep_cache_sz; i++) {
 /* Skip the closed sockets */
@@ -2207,7 +2212,7 @@
 i, elapsed, (int)(difftime(time(NULL), 
rt)));
 }
 }
-if (n > aw->ep_mincache_sz) {
+if ((cnt - n) < aw->ep_mincache_sz) {
 if (JK_IS_DEBUG_LEVEL(l)) {
 jk_log(l, JK_LOG_DEBUG,
 "reached pool min size %u from %u cache slots",



-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Re: NIO vs BIO speed

2006-06-26 Thread devlists

I'm not seeing the windows problems, I used to have many problems on
windows using jdk1.4, but since 1.5 I have no problems at all. I do have a
brand new machine, so maybe there is some windows patch on it that I
didn't have before, other than that I can't think of anything.

I'm showing almost identical results between NIO and blocking IO on both
windows and FC4.

here is an output from my test run on windows, 100,000 requests on each
connector

I still have one thing to improve, to use an java.util.concurrent
Exchanger  on the read that registers back with the Selector and handles
the hand off between two threads correctly.

C:\development\covalent\ers\apache2.0\bin>java -version
java version "1.5.0_06"
Java(TM) 2 Runtime Environment, Standard Edition (build 1.5.0_06-b05)
Java HotSpot(TM) Client VM (build 1.5.0_06-b05, mixed mode, sharing)

--from tomcat startup, two connectors starting up,.
Jun 26, 2006 3:16:53 PM org.apache.coyote.http11.Http11Protocol start
INFO: Starting Coyote HTTP/1.1 on http-8080
Jun 26, 2006 3:16:53 PM org.apache.coyote.http11.Http11NioProtocol start
INFO: Starting Coyote HTTP/1.1 on http-8081
Jun 26, 2006 3:16:53 PM org.apache.catalina.startup.Catalina start
INFO: Server startup in 1828 ms


C:\development\covalent\ers\apache2.0\bin>ab -n 10 -c 20 -k 
http://127.0.0.1:8081/tomcat.gif
This is ApacheBench, Version 2.0.41-dev <$Revision: 1.2 $> apache-2.0
Copyright (c) 1996 Adam Twiss, Zeus Technology Ltd, http://www.zeustech.net/
Copyright (c) 1998-2002 The Apache Software Foundation,
http://www.apache.org/

Benchmarking 127.0.0.1 (be patient)
Completed 1 requests
Completed 2 requests
Completed 3 requests
Completed 4 requests
Completed 5 requests
Completed 6 requests
Completed 7 requests
Completed 8 requests
Completed 9 requests
Finished 10 requests


Server Software:Apache-Coyote/1.1
Server Hostname:127.0.0.1
Server Port:8081

Document Path:  /tomcat.gif
Document Length:1934 bytes

Concurrency Level:  20
Time taken for tests:   16.734375 seconds
Complete requests:  10
Failed requests:0
Write errors:   0
Keep-Alive requests:10
Total transferred:  21640 bytes
HTML transferred:   19340 bytes
Requests per second:5975.72 [#/sec] (mean)
Time per request:   3.347 [ms] (mean)
Time per request:   0.167 [ms] (mean, across all concurrent requests)
Transfer rate:  12628.38 [Kbytes/sec] received

Connection Times (ms)
  min  mean[+/-sd] median   max
Connect:00   0.0  0   0
Processing: 03   6.3  0  46
Waiting:03   6.3  0  46
Total:  03   6.3  0  46

Percentage of the requests served within a certain time (ms)
  50%  0
  66%  0
  75%  0
  80% 15
  90% 15
  95% 15
  98% 15
  99% 15
 100% 46 (longest request)

C:\development\covalent\ers\apache2.0\bin>ab -n 10 -c 20 -k 
http://127.0.0.1:8080/tomcat.gif
This is ApacheBench, Version 2.0.41-dev <$Revision: 1.2 $> apache-2.0
Copyright (c) 1996 Adam Twiss, Zeus Technology Ltd, http://www.zeustech.net/
Copyright (c) 1998-2002 The Apache Software Foundation,
http://www.apache.org/

Benchmarking 127.0.0.1 (be patient)
Completed 1 requests
Completed 2 requests
Completed 3 requests
Completed 4 requests
Completed 5 requests
Completed 6 requests
Completed 7 requests
Completed 8 requests
Completed 9 requests
Finished 10 requests


Server Software:Apache-Coyote/1.1
Server Hostname:127.0.0.1
Server Port:8080

Document Path:  /tomcat.gif
Document Length:1934 bytes

Concurrency Level:  20
Time taken for tests:   16.937500 seconds
Complete requests:  10
Failed requests:0
Write errors:   0
Keep-Alive requests:98218
Total transferred:  216397582 bytes
HTML transferred:   193405802 bytes
Requests per second:5904.06 [#/sec] (mean)
Time per request:   3.388 [ms] (mean)
Time per request:   0.169 [ms] (mean, across all concurrent requests)
Transfer rate:  12476.75 [Kbytes/sec] received

Connection Times (ms)
  min  mean[+/-sd] median   max
Connect:00   0.3  0  15
Processing: 03  16.8  0 343
Waiting:03  16.7  0 343
Total:  03  16.9  0 343

Percentage of the requests served within a certain time (ms)
  50%  0
  66%  0
  75%  0
  80%  0
  90% 15
  95% 15
  98% 15
  99% 62
 100%343 (longest request)

C:\development\covalent\ers\apache2.0\bin>

> I haven't tested Filip's connector yet, but that's also my experience with
> the AJP/NIO connector:  NIO is pretty much useless on Windows.

> > -Original Message-
> > From: Remy Maucherat [mailto:[EMAIL PROTECTED]
> > Sent: Monday, June 26, 2006 4:58 AM
> > To: Tomcat Developers

Re: NIO vs BIO speed


[EMAIL PROTECTED] wrote:

I'm not seeing the windows problems, I used to have many problems on
windows using jdk1.4, but since 1.5 I have no problems at all. I do have a
brand new machine, so maybe there is some windows patch on it that I
didn't have before, other than that I can't think of anything.



Did you ever tried using 500+ concurrent connections.
With < 200 BIO is just fine on any machine with 512MB RAM.

Try to play with larger numbers like few thousands, and
then you'll see the real stuff.

Regards,
Mladen.

-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

svn commit: r417296 - /tomcat/connectors/trunk/jk/native/common/jk_status.c

2006-06-26 Thread rjung

Author: rjung
Date: Mon Jun 26 14:07:14 2006
New Revision: 417296

URL: http://svn.apache.org/viewvc?rev=417296&view=rev
Log:
Tried to fix some HTML syntax errors in output.
Most of it: Thanks to Johan Bergstroem.

Modified:
tomcat/connectors/trunk/jk/native/common/jk_status.c

Modified: tomcat/connectors/trunk/jk/native/common/jk_status.c
URL: 
http://svn.apache.org/viewvc/tomcat/connectors/trunk/jk/native/common/jk_status.c?rev=417296&r1=417295&r2=417296&view=diff
==
--- tomcat/connectors/trunk/jk/native/common/jk_status.c (original)
+++ tomcat/connectors/trunk/jk/native/common/jk_status.c Mon Jun 26 14:07:14 
2006
@@ -511,19 +511,19 @@
 wr->s->jvm_route, ")\n", NULL);
 jk_putv(s, "req_uri, "\">\n", NULL);
-jk_puts(s, "\n\n");
 jk_puts(s, "s->name, "\">\n", NULL);
 jk_puts(s, "\n\n", selected);
+jk_printf(s, "value=\"%u\">\n", selected);
 jk_puts(s, "\n\n", i);
+jk_printf(s, "value=\"%u\">\n", i);
 
 jk_puts(s, "\nDistance:\n", wr->s->distance);
+jk_printf(s, "value=\"%d\"/>\n", wr->s->distance);
 jk_puts(s, "Load factor:\n", wr->s->lb_factor);
+jk_printf(s, "value=\"%d\"/>\n", wr->s->lb_factor);
 jk_puts(s, "Route Redirect:s->redirect, NULL);
 jk_puts(s, "\"/>\n");
@@ -538,7 +538,7 @@
 if (wr->s->is_stopped)
 jk_puts(s, "  checked=\"checked\"");
 jk_puts(s, "/>\n");
-jk_puts(s, "\n\n");
+jk_puts(s, "\n");
 jk_puts(s, "\n\n");
 
 }
@@ -548,12 +548,12 @@
 dworker, "\n", NULL);
 jk_putv(s, "req_uri, "\">\n", NULL);
-jk_puts(s, "\n\n");
 jk_puts(s, "\n", NULL);
 jk_puts(s, "\n\n", i);
+jk_printf(s, "value=\"%u\"/>\n", i);
 
 jk_puts(s, "\nRetries:\n", lb->s->retries);
@@ -578,7 +578,6 @@
 "TypeHostAddr"
 "\n");
 jk_putv(s, "", status_worker_type(w->type), "", NULL);
-jk_puts(s, "\n\n");
 jk_printf(s, "%s:%d", aw->host, aw->port);
 jk_putv(s, "", jk_dump_hinfo(&aw->worker_inet_addr, buf),
 "\n\n", NULL);
@@ -782,7 +781,7 @@
 i = status_int("wf", s->query_string, wr->s->lb_factor);
 if (i > 0 && wr->s->lb_factor != i) {
 wr->s->lb_factor = i;
-/* Recalculate the load multiplicators wrt. lb_factor */
+/* Recalculate the load multiplicators wrt. lb_factor */
 update_mult(lb, l);
 }
 }



-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Re: Jk_status doesn't generate valid html (patch included)

Thanks, I committed most of it (and some more). The only thing I left 
unchanged is not escaping the Ampersand as a seperator between query 
arguments.


Johan Bergström schrieb:

Hello dev-list.
I'm quite new to both tomcat and mod_jk.. really liking it so far! 
Thought I could contribute back with some minor things I found.


Jk_status throws out some garbled html;

 * Ampersands (&) are not escaped
 * Input hiddens are wrapped in a table (?)
 * Superfluous td/tr/table and invalid closures scattered throughout 
the file.

 and
 * (Tabbed in an comment to follow guidelines)

Maybe i missed something (like a reason why they should be there) but 
here's an proposed diff:


(have a nice weekend)




-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

svn commit: r417297 - /tomcat/connectors/trunk/jk/xdocs/changelog.xml

2006-06-26 Thread rjung

Author: rjung
Date: Mon Jun 26 14:10:57 2006
New Revision: 417297

URL: http://svn.apache.org/viewvc?rev=417297&view=rev
Log:
Adding most of my changes to the changelog.

Modified:
tomcat/connectors/trunk/jk/xdocs/changelog.xml

Modified: tomcat/connectors/trunk/jk/xdocs/changelog.xml
URL: 
http://svn.apache.org/viewvc/tomcat/connectors/trunk/jk/xdocs/changelog.xml?rev=417297&r1=417296&r2=417297&view=diff
==
--- tomcat/connectors/trunk/jk/xdocs/changelog.xml (original)
+++ tomcat/connectors/trunk/jk/xdocs/changelog.xml Mon Jun 26 14:10:57 2006
@@ -34,13 +34,64 @@
   the connection if client connection is broken during the request. (mturk)
   
   
-  Renamed cache_timeout firective to connection_pool_timeout. (mturk)
+  Renamed cache_timeout directive to connection_pool_timeout. (mturk)
   
   
   Added connection_pool_minsize directive. (mturk)
   
   
   Deprecate recycle_timeout directive. (mturk)
+  
+  
+  Corrected some HTML syntax bugs in output of status worker. (rjung)
+  
+  
+  Added the refresh=n parameter to the status worker. It will update the 
display every n seconds. (rjung)
+  
+  
+  Balancer: Add attribute distance to balanced workers to express 
preferences between workers. (rjung)
+  
+  
+  Balancer: Add attribute jvm_route to balanced workers to be able to use 
the same target in different balancers. (rjung)
+  
+  
+  Status: Add lb_mult to status. (rjung)
+  
+  
+  Balancer: Make different balancing strategies work in a similar way (use 
lb_value, use decay during global maintenance, use integer factors for weights. 
(rjung)
+  
+  
+  Balancer: Improve locking. (rjung)
+  
+  
+  Balancer: Workers start slower after recovering. (rjung)
+  
+  
+  Balancer: Make different balancing strategies work in a similar way (use 
lb_value, use decay during global maintenance, use integer factors lb_mult for 
weights). (rjung)
+  
+  
+  Balancer: Move recovery check to global maintenance. (rjung)
+  
+  
+  Balancer: Add global maintenance method, that is called in only one 
process. (rjung)
+  
+  
+  Extend our use of autoconf to find a 32Bit and a 64Bit unsigned type and 
their printf formats. (rjung)
+  
+  
+  Logging: piped loggers for JkLogFile and Apache 1.3. (rjung)
+  
+  
+  Logging: Add PID to log lines for each log level apart from REQUEST. 
(rjung)
+  
+  
+  Logging: flush buffered logs to keep lines in correct order. Output 
final newline together with log message. (rjung)
+  
+  
+  Reducing shm size. (rjung)
+  
+  
+  Only log removing of old worker, when we actually do it. (rjung)
   
   
   37469: Fix shared memory close for forked childs.



-
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Re: svn commit: r417288 - /tomcat/connectors/trunk/jk/native/common/jk_ajp_common.c