HTTP Bind problems

I’m having serious issues with getting my iJab client to work 100% with openfire.

iJab connect and it looks like its working fine, but when we look at the logs (and some console logging on the client side), I’m getting alot of errors coming through.

The problem seems to come about once all the connection stuff is done and the client sends through the initial XML payload packet (taken from the nohup log once log.httpbind.enabled set to true):

Fri May 13 14:26:53 SAST 2011: HTTP RECV(de63534c):

At this point, the server just sits there and does nothing, then after about 30-40 seconds, I get a warning in the warning log

2011.05.13 14:27:23 Error parsing user request. [192.168.2.1]

Then immediately after that I get the anotehr 3 payload packets coming through as well as 3 warning messages

nohup:

Fri May 13 14:27:23 SAST 2011: HTTP RECV(de63534c):

Fri May 13 14:27:23 SAST 2011: HTTP RECV(de63534c):

Fri May 13 14:27:23 SAST 2011: HTTP RECV(de63534c):

warn:

2011.05.13 14:27:23 Client provided invalid session: de63534c. [192.168.2.1]

2011.05.13 14:27:23 Client provided invalid session: de63534c. [192.168.2.1]

2011.05.13 14:27:23 Client provided invalid session: de63534c. [192.168.2.1]

I’ve done some investigation and from what I can see, the server seems to be timing out on something, I’m not sure what, but it’s then closing the session and then recieves more packets, which are 3 queued up payload packets. Since the session has now been closed, then cause the invalid session error.

On the client side, I get a “bad request” error for the first post and then 3 “invalid SID” errors. After all this happens, the connection just get re-established.

I would really appreciate some help with this. I’ve gone through the code, but to be honest, I get a little lost after a while. If anyone could point me in the right direction, I might be able to fix the problem myself.

Hello,

I feel your pain A couple of things. Remember that BOSH is long polling, so after sending something from the client, the server will ‘hold’ the request until a timeout is reached and then return to the client. You may wish to check this setting:

xmpp.httpbind.client.requests.ignoreOveractivity

in the openfire source code. Setting it to true may help you some.

daryl

I set the Jive Global in the system properties to true as you suggested, but this didn’t help at all. It’s still holding up and timing out on the payload packet.

Anyone else have any insight?