[Red5] red5 server -> complete halt

[Daniel Daley]

I think he just meant that the new keep-alive code that disconnects  
ghost connections should also take care of live connections. So the  
problem we are having with a stream getting stuck and returning  
BadName shouldn't be occurring due to a ghost connection.

Thanks,

--Dan--


On Dec 9, 2006, at 1:44 PM, Adam wrote:

> Hi Steven,
>
> could you please be so gentle and explaine to you me what do mean by:
>
> >Well, there's no specific difference between a connection for VOD  
> and a
> >connection for Live.
>
> iam not getting your point!
>
> thanks
> -Adam-
>
> >
> > Hi List,
> >
> > ived finished testing on rtmpMinaConnection = 10000,  
> rtmptConnection =
> > 5000 with more than 100 Clients. The good first.
> > Red5 isnt halting/freezing for this test with the above mentioned  
> values.
> > BUT,
> > after about 20h the publisher got diconnected. i guess this ghost
> > connection removel routine killed also the publisher.
> >
> Well, there's no specific difference between a connection for VOD  
> and a
> connection for Live.
> A publisher reconnect leads to badname status.
> > i ve closed all clients. i even shutdown the machines executing  
> the test
> > clients.
> > but on red5 side 3 connection remained.
> > i guess this goes along what Daniel Daley mentioned in his thread:
> > changeset 1582 ping/pong keepalive.
> >
> Ok, I will look into this problem.
> This behavior isnt a big deal. I guess i can solve it with the  
> technics ive
> > mentioned in the thread: Cannot subscribe to Live Stream if its  
> being
> > Recorded. (avoiding badname)
> >
> > ill keep on testing with other values. hopefully i can find a
> > configuration where i dont need any detours.
> >
> > greets,
> > -Adam-
> _______________________________________________
> Red5 mailing list
> Red5 at osflash.org
> http://osflash.org/mailman/listinfo/red5_osflash.org

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://osflash.org/pipermail/red5_osflash.org/attachments/20061209/28fa64c6/attachment-0001.htm