It seemed the problem was fixed. However, as we set rtmpt as
the only channel, our other users were experiencing a big drop off
in performance due to the tunneling.
I read a very good blog entry:
http://www.infoaccelerator.net/blog/post.cfm/setting-up-rtmpt-failover-on-lcdswhere Andrew talks about using rtmpt as failover for those
users that cannot use rtmp.
We implemented the structure he described:
- channel definition for rtmp in services-config.xml
- channel definition for rtmpt in services-config.xml
- channel definition for polling-amf in services-config.xml
- define both channels as default channels in
data-management-config.xml in the order rtmp, rtmpt, polling-amf
Everybody was able to use the data services..