You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@tomcat.apache.org by ma...@apache.org on 2010/09/30 18:16:05 UTC
svn commit: r1003136 -
/tomcat/trunk/java/org/apache/coyote/ajp/AjpAprProtocol.java
Author: markt
Date: Thu Sep 30 16:16:05 2010
New Revision: 1003136
URL: http://svn.apache.org/viewvc?rev=1003136&view=rev
Log:
Fix remaining issues with AJP/APR and TCK. Note mod_jk log is full of warnings so there is still work to do here.
Modified:
tomcat/trunk/java/org/apache/coyote/ajp/AjpAprProtocol.java
Modified: tomcat/trunk/java/org/apache/coyote/ajp/AjpAprProtocol.java
URL: http://svn.apache.org/viewvc/tomcat/trunk/java/org/apache/coyote/ajp/AjpAprProtocol.java?rev=1003136&r1=1003135&r2=1003136&view=diff
==============================================================================
--- tomcat/trunk/java/org/apache/coyote/ajp/AjpAprProtocol.java (original)
+++ tomcat/trunk/java/org/apache/coyote/ajp/AjpAprProtocol.java Thu Sep 30 16:16:05 2010
@@ -481,9 +481,9 @@ public class AjpAprProtocol
if (state != SocketState.LONG && state != SocketState.ASYNC_END) {
connections.remove(socket);
recycledProcessors.offer(result);
- if (state == SocketState.OPEN) {
- proto.endpoint.getPoller().add(socket.getSocket().longValue());
- }
+ //if (state == SocketState.OPEN) {
+ // proto.endpoint.getPoller().add(socket.getSocket().longValue());
+ //}
}
}
}
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@tomcat.apache.org
For additional commands, e-mail: dev-help@tomcat.apache.org
Re: svn commit: r1003136 - /tomcat/trunk/java/org/apache/coyote/ajp/AjpAprProtocol.java
Posted by Rainer Jung <ra...@kippdata.de>.
On 30.09.2010 19:10, Mark Thomas wrote:
> On 30/09/2010 17:46, Mark Thomas wrote:
>> I took a look with Wireshark and it appears the END_RESPONSE message is
>> being sent in the same packet as the response body. For the AJP/BIO
>> connector they are in separate packets. Not sure what, if anything, that
>> has to do with the problem.
>
> I think I have found the problem - too much re-factoring of the flush()
> method. Just running the TCK now...
Quite possible. If it doesn't fix it and you can reproduce easily, tur
JkLogLevel to trace and reproduce. That should show us, at what point
mod_jk is missing something in the response.
Regards,
Rainer
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@tomcat.apache.org
For additional commands, e-mail: dev-help@tomcat.apache.org
Re: svn commit: r1003136 - /tomcat/trunk/java/org/apache/coyote/ajp/AjpAprProtocol.java
Posted by Mark Thomas <ma...@apache.org>.
On 30/09/2010 17:46, Mark Thomas wrote:
> I took a look with Wireshark and it appears the END_RESPONSE message is
> being sent in the same packet as the response body. For the AJP/BIO
> connector they are in separate packets. Not sure what, if anything, that
> has to do with the problem.
I think I have found the problem - too much re-factoring of the flush()
method. Just running the TCK now...
Mark
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@tomcat.apache.org
For additional commands, e-mail: dev-help@tomcat.apache.org
Re: svn commit: r1003136 - /tomcat/trunk/java/org/apache/coyote/ajp/AjpAprProtocol.java
Posted by Mark Thomas <ma...@apache.org>.
On 30/09/2010 17:28, Rainer Jung wrote:
> On 30.09.2010 18:16, markt@apache.org wrote:
>> Author: markt
>> Date: Thu Sep 30 16:16:05 2010
>> New Revision: 1003136
>>
>> URL: http://svn.apache.org/viewvc?rev=1003136&view=rev
>> Log:
>> Fix remaining issues with AJP/APR and TCK. Note mod_jk log is full of
>> warnings so there is still work to do here.
>
> If you like, I can have a look at them. Just send me the jk log and the
> version info of mod_jk you are using. JkLogLevel at least "info" would
> be helpful.
That would be great. I'm sure my re-factoring has broken something.
Every request is generating the following in the mod_jk log:
[Thu Sep 30 17:40:53.765 2010] [4460:4444] [info] jk_ajp_common.c
(2540): (worker1) sending request to tomcat failed (recoverable),
because of server error (attempt=1)
[Thu Sep 30 17:40:55.875 2010] [4460:4444] [info] jk_ajp_common.c
(1490): (worker1) did not receive END_RESPONSE, closing socket -1
I took a look with Wireshark and it appears the END_RESPONSE message is
being sent in the same packet as the response body. For the AJP/BIO
connector they are in separate packets. Not sure what, if anything, that
has to do with the problem.
Mark
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@tomcat.apache.org
For additional commands, e-mail: dev-help@tomcat.apache.org
Re: svn commit: r1003136 - /tomcat/trunk/java/org/apache/coyote/ajp/AjpAprProtocol.java
Posted by Rainer Jung <ra...@kippdata.de>.
On 30.09.2010 18:16, markt@apache.org wrote:
> Author: markt
> Date: Thu Sep 30 16:16:05 2010
> New Revision: 1003136
>
> URL: http://svn.apache.org/viewvc?rev=1003136&view=rev
> Log:
> Fix remaining issues with AJP/APR and TCK. Note mod_jk log is full of warnings so there is still work to do here.
If you like, I can have a look at them. Just send me the jk log and the
version info of mod_jk you are using. JkLogLevel at least "info" would
be helpful.
Note: you can ignore everything that is "[info]" or lower. Only "[warn]"
and "[error]" really matter.
Regards,
Rainer
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@tomcat.apache.org
For additional commands, e-mail: dev-help@tomcat.apache.org