Are you 100% sure that you are clearing the intervals on the server? Or are you setting intervals more then once, like each time a client connects?
The interval is set in an zone level extension and cleared on destroy()
You can test this by calling a trace command every time an interval is called and looking at the logs.
Code: Select all
INFO | jvm 1 | 2007/04/03 15:37:38 | [extUserLogin.as]: >> INTERVAL SET
I traced it as requested and its only invoking once ..
To be safe you should always call clearInterval before using setInterval.
i have set the extention to clear any existing interval before getting the new one. infact there is only one global interval set for the entire app.
If you reset the server does this happen right away, or only after the server has been running for a while?
No this happens right away without any delay ..... even the very next moment the user connects
In order to see if it's the server kicking the users (I don't think it is) you should inspect the logs and see if you find messeges like this:
Code:
INFO: Disconneting idle user: {USERNAME}
The server logs (with INFO level) all users that are being kicked because of inactivity
Code: Select all
INFO | jvm 1 | 2007/04/03 15:45:27 | 15:45:27.234 - [ WARNING ] > Extension Call has invalid RoomID. User = anand - /59.162.126.220
INFO | jvm 1 | 2007/04/03 15:45:27 | 15:45:27.359 - [ WARNING ] > Extension Call has invalid RoomID. User = anand - /59.162.126.220
INFO | jvm 1 | 2007/04/03 15:45:27 | 15:45:27.359 - [ WARNING ] > Extension Call has invalid RoomID. User = anand - /59.162.126.220
INFO | jvm 1 | 2007/04/03 15:45:39 | 15:45:39.265 - [ INFO ] > Disconneting idle user: anand
These are the lines that was logged which was not an usual behavior.
What type of server / hosting are you using? Is it dedicated? How much bandwidth do you use?
Code: Select all
INFO | jvm 1 | 2007/04/03 15:37:38 | --- [ System Info ] ------------------------------------------
INFO | jvm 1 | 2007/04/03 15:37:38 |
INFO | jvm 1 | 2007/04/03 15:37:38 | System CPU(s): 2
INFO | jvm 1 | 2007/04/03 15:37:38 | VM Max memory: 66 MB
INFO | jvm 1 | 2007/04/03 15:37:38 |
INFO | jvm 1 | 2007/04/03 15:37:38 | os.name: Windows 2003
INFO | jvm 1 | 2007/04/03 15:37:38 | os.arch: x86
INFO | jvm 1 | 2007/04/03 15:37:38 | os.version: 5.2
INFO | jvm 1 | 2007/04/03 15:37:38 | java.version: 1.5.0_11
INFO | jvm 1 | 2007/04/03 15:37:38 | java.vendor: Sun Microsystems Inc.
INFO | jvm 1 | 2007/04/03 15:37:38 | java.vendor.url: http://java.sun.com/
INFO | jvm 1 | 2007/04/03 15:37:38 | java.vm.specification.version: 1.0
INFO | jvm 1 | 2007/04/03 15:37:38 | java.vm.version: 1.5.0_11-b03
INFO | jvm 1 | 2007/04/03 15:37:38 | java.vm.vendor: Sun Microsystems Inc.
INFO | jvm 1 | 2007/04/03 15:37:38 | java.vm.name: Java HotSpot(TM) Client VM
INFO | jvm 1 | 2007/04/03 15:37:38 |
yes the server is run on an 200KBps Shared Web server.
and can u guess why the strange behavior ?? while lapo tested ?? is it all connected ... i think the plot not loading issue is with the asynchronous nature of AS. is there a workaround ?? and is there a solution to my main problem ??
UPDATE*
Firewall used in our server is visnetic
Please help me ..... i am going mad with this behavior of SFS on my server ... the strange thing is that i have the same app working fine on my developer machine ...
