Dear all,
Am still having this problem, so if anyone has any ideas, please let me
know.
Here is some more information.....
The client and Tomcat JVM complete the handshake. Then there are three
netbios or nbstat lookups, each of which seems to time out after 1.5
seconds.
This accounts for the 4.5 seconds between the handshake and the
POST data being sent.
What is strange is that several other sockets to Tomcat, which are
performing GETs, work very well, with no delays.
So is it possible that with the POST, Tomcat is causing Windows to
do a NBSTAT or NETBIOS lookup?
Any ideas on what the problem is or how to further diagnose it?
Thanks,
Tim
----- Original Message -----
From: "Tim Lavers" <[EMAIL PROTECTED]>
To: "Tomcat Users List" <users@tomcat.apache.org>
Sent: Wednesday, April 16, 2008 3:13 PM
Subject: Re: Delay between TCP handshake and POST (complete)
Just checked: our firewall does not do DNS lookups.
Tim
----- Original Message -----
From: "Ronald Klop" <[EMAIL PROTECTED]>
To: "Tomcat Users List" <users@tomcat.apache.org>
Cc: "Tim Lavers" <[EMAIL PROTECTED]>
Sent: Tuesday, April 15, 2008 8:58 PM
Subject: Re: Delay between TCP handshake and POST (complete)
Do you have a firewall which does DNS lookups?
Ronald.
On Tue Apr 15 08:58:18 CEST 2008 Tomcat Users List
<users@tomcat.apache.org> wrote:
Thanks Ronald,
I've got enableLookups="false" in the server.xml file
and Wireshark shows no signs of DNS lookups.
Also, all of the Webstart stuff (downloading jars etc)
is pretty responsive.
So I don't think that is the problem ... but is there any way
that Tomcat would do reverse DNS lookups if
I've got enableLookups=false"?
Tim
> Reverse DNS lookup which always fails?
>
> Ronald.
>
> On Mon Apr 14 23:46:23 CEST 2008 Tomcat Users List
> <users@tomcat.apache.org> wrote:
>> Am running a Tomcat 6.0 server that is distributing a client program
>> using
>> Java WebStart. The client program communicates to the server using
>> an RMI HTTP tunnelling servlet which is running inside the same
>> Tomcat
>> instance. I've had this setup running on three different computers.
>> On two of them, the performance is pretty good. On our production
>> system, however, the performance is terrible.
>>
>> The download of the application is ok, what is bad is a strange pause
>> in the tunnelling. Using wireshark, I can see that after the WebStart
>> system has downloaded all of the software, the tunnelling operation
>> initiates a TCP connection with the three-way handshake but then
>> there
>> is a 4.5 second delay before the expected POST operation.
>>
>> The differences between the slow system and the two fast systems are:
>> 1. The slow one has two network cards
>> 2. The slow one is not a Domain Controller, its domain controller
>> is a Thinpoint Linux machine.
>>
>> The delay is always very close to 4.5 seconds, so perhaps it is a
>> timeout.
>> Anyhow, any advice on how to further diagnose the problem would be
>> much appreciated.
>>
>> ---------------------------------------------------------------------
>> To start a new topic, e-mail: users@tomcat.apache.org
>> To unsubscribe, e-mail: [EMAIL PROTECTED]
>> For additional commands, e-mail: [EMAIL PROTECTED]
>>
>
>
---------------------------------------------------------------------
To start a new topic, e-mail: users@tomcat.apache.org
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
---------------------------------------------------------------------
To start a new topic, e-mail: users@tomcat.apache.org
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
---------------------------------------------------------------------
To start a new topic, e-mail: users@tomcat.apache.org
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
