FMS Recording Authentication

Question

Hi all,
I'm hoping an advanced FMS user/administrator (or Adobe)
might be able to offer some advice here. My environment is this:
- One server runs an administrative publishing system for
(e.g) joewebsite.com. Authors (authenticated users) are able to
post content to joewebsite.com. They are also able to record their
webcam/microphone to create podcasts and videocasts.
- A second server resides in another location (not LAN) and
runs FMS in a LAMP environment, to receive authors video/audio and
then move it to a web-accessible folder for user download (not
using a Flash Streaming server, just progressive download).
What I'd like to do is this: From the SWF recorder a web
request (i.e. not rtmp/e) is sent to the FMS server with a session
ID. The server checks to make sure that the session is valid as per
the publishing system. Assuming it's valid, the system tells the
FMS to allow that user's IP to connect to the FMS. The system then
sends back to the user's SWF the filename that should be used for
recording as per NetStream.publish(fileName, "record"), so that the
system can keep track of what file is used for which recording.
After the user is done recording (i.e. they hit the Stop button),
the SWF sends an "I'm done" to the server, and the file is moved
from its FMS location to the web-accessible folder.
The main thing is this: a non-firewall solution to allowing
only certain IPs to connect to the FMS. I've looking into the
Users.xml, but that unfortunately requires an FMS reboot. As the
system could be recording more than one stream at the same time,
this isn't a viable solution.
FWIW: I'm working on a Developer's license for FMS 3, and the
company I work for is pretty serious about buying an Interactive
Server license if I can get this thing working.
I appreciate any time and advice. I'm a big fan of reading
manuals and don't need my hand held, so if you could even just
point me in the right direction with a page number, I could figure
out the rest.
Regards,
N

Answer 1

Are you using Stream.flush API - if you are using that it migth cause some syncronisation issues if data and audio/video is mixed.
I did not get your usecase clearly - let me know if you using above API. I think it would be better to use second option if you can usedata and audio seperately but probably you can explain use case better so that we migth help you.