RE: Error? org/apache/ajp/tomcat4/Ajp13Connector createProcessor()

2002-12-02 Thread Kevin Seguin
seems like a bug.  

either way, I believe the org.apache.ajp package is deprecated.

 -Original Message-
 From: eric scroger [mailto:[EMAIL PROTECTED]] 
 Sent: Monday, December 02, 2002 12:40 PM
 To: Tomcat Developers List
 Subject: Error? org/apache/ajp/tomcat4/Ajp13Connector 
 createProcessor() 
 
 
 Greetings,
 
 We were noticing some 'No processor available' exceptions
 associated with the Ajp13Connectors under heavy load.
 I downloaded the Tomcat 4.0.6 connectors source and had
 a look at the code responsible for creating a new processor to
 handle an incoming Ajp13 socket connection from Apache.
 
 In org.apache.ajp.tomcat4.Ajp13Connector.java, in the
 createProcessor() method, I noticed that there was no logic
 to handle the case if maxProcessors was configured to -1
 for infinite processors.  In the HttpConnector source, if
 maxProcessors was -1, then it returned a newProcessor().
 I modified this createProcessor() method to function in the
 same manner and the Ajp13 No processor available
 exceptions disappeared.
 
   private Ajp13Processor createProcessor() {
 
   synchronized (processors) {
   if (processors.size()  0)
   return ((Ajp13Processor) processors.pop());
   if ((maxProcessors  0)  (curProcessors  maxProcessors))
   return (newProcessor());
   else {
   if ((maxProcessors  0))// my new code 
 returns a new 
 processor
   return (newProcessor());// if maxProcessors 
 is -1, 
 aka infinite.
  else   return (null);

 // original else return result
   }
   }
 }
 
 My question then, is this a bug in the Ajp13Connector functionality?
 Or is there some design reason to not allow the user to configure
 the Ajp13Connector for infinite processors via the -1 option.
 I couldn't locate any documentation that says Ajp13 connectors
 cannot and should not be configured for infinite processors.
 
 Thanks,
 
 Eric Scroger
 
 
 
 
 
 --
 To unsubscribe, e-mail:   
 mailto:tomcat-dev- [EMAIL PROTECTED]
 For 
 additional commands, 
 e-mail: mailto:[EMAIL PROTECTED]
 
 

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




Re: Error? org/apache/ajp/tomcat4/Ajp13Connector createProcessor()

2002-12-02 Thread eric scroger
Thanks.

I figured it was a bug.  Unless anyone has any other information to the 
contrary,
I'm going to file a bug report today.  

Seeing the Ajp package is deprecated, are the Warp connectors fairly stable?
The documentation for the Warp connectors is weak (FIXME's) and incomplete
when compared to the Ajp documentation.

Eric

Kevin Seguin wrote:

seems like a bug.  

either way, I believe the org.apache.ajp package is deprecated.

 

-Original Message-
From: eric scroger [mailto:[EMAIL PROTECTED]] 
Sent: Monday, December 02, 2002 12:40 PM
To: Tomcat Developers List
Subject: Error? org/apache/ajp/tomcat4/Ajp13Connector 
createProcessor() 


Greetings,

We were noticing some 'No processor available' exceptions
associated with the Ajp13Connectors under heavy load.
I downloaded the Tomcat 4.0.6 connectors source and had
a look at the code responsible for creating a new processor to
handle an incoming Ajp13 socket connection from Apache.

In org.apache.ajp.tomcat4.Ajp13Connector.java, in the
createProcessor() method, I noticed that there was no logic
to handle the case if maxProcessors was configured to -1
for infinite processors.  In the HttpConnector source, if
maxProcessors was -1, then it returned a newProcessor().
I modified this createProcessor() method to function in the
same manner and the Ajp13 No processor available
exceptions disappeared.

 private Ajp13Processor createProcessor() {

 synchronized (processors) {
 if (processors.size()  0)
 return ((Ajp13Processor) processors.pop());
 if ((maxProcessors  0)  (curProcessors  maxProcessors))
 return (newProcessor());
 else {
 if ((maxProcessors  0))// my new code 
returns a new 
processor
 return (newProcessor());// if maxProcessors 
is -1, 
aka infinite.
else   return (null);
  
// original else return result
 }
 }
}

My question then, is this a bug in the Ajp13Connector functionality?
Or is there some design reason to not allow the user to configure
the Ajp13Connector for infinite processors via the -1 option.
I couldn't locate any documentation that says Ajp13 connectors
cannot and should not be configured for infinite processors.

Thanks,

Eric Scroger





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


   


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

 




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