I currently have 3 mountpoints, 128, 56 and 24. I'm using darkice as the source and I'm relaying to another server running icecast2. I'm assuming that darkice is using lame for transcoding the bit-rates on the fly and sending 128k/s + 56k/s + 24k/s to the relay host... which is using a lot of bandwidth on the source network (which I don't have much to spare). Is there any way to just send one 128k/s mountpoint and have the relay side transcode to the bitrates and go from there, that way I'm saving 80k/s of upload from the source host. The stream randomly starts to glitch out every once and awhile and I'm trying to fix it. Any pointers would be greatly appreciated. -- Cody Tubbs : (Certified Master Linux Administrator) : (Certified Unix Security Specialist) -- wISPdirect : http://www.wISPdirect.com -- Broadway Internet
On Fri, 2004-12-10 at 16:32, Cody Tubbs (wISPdirect) wrote:> I currently have 3 mountpoints, 128, 56 and 24. I'm using darkice as > the source and I'm relaying to another server running icecast2. I'm > assuming that darkice is using lame for transcoding the bit-rates on > the fly and sending 128k/s + 56k/s + 24k/s to the relay host... which > is using a lot of bandwidth on the source network (which I don't have > much to spare). Is there any way to just send one 128k/s > mountpoint and have the relay side transcode to the bitrates and go > from there, that way I'm saving 80k/s of upload from the source > host. > The stream randomly starts to glitch out every once and awhile and > I'm trying to fix it. Any pointers would be greatly appreciated.sure, sending the 128k stream works as normal and just have 2 local transcoders to take the 128k stream from the relay, reencode and push back to the relay under a different mountpoint. The likes of streamtranscoder can do that, but I believe you can use a "curl | ices" type solution as well. Not sure about darkice but it may do. karl
Ah so basically you mean catch the 128k stream locally on the relay server, and re-source it to localhost while transcoding to the other two bitrates? Basically the 56k and 24k mounts will be delayed more than the 128 though I'm assuming (which isnt really a problem). I don't know if the above is what you meant though, but I think it will work, I havn't checked out streamtranscoder yet. Thanks. -- Cody Tubbs : (Certified Master Linux Administrator) : (Certified Unix Security Specialist) -- wISPdirect : http://www.wISPdirect.com -- Broadway Internet On 10 Dec 2004 at 16:46, Karl Heyes wrote:> On Fri, 2004-12-10 at 16:32, Cody Tubbs (wISPdirect) wrote: > > I currently have 3 mountpoints, 128, 56 and 24. I'm using darkice as > > the source and I'm relaying to another server running icecast2. I'm > > assuming that darkice is using lame for transcoding the bit-rates on > > the fly and sending 128k/s + 56k/s + 24k/s to the relay host... which > > is using a lot of bandwidth on the source network (which I don't have > > much to spare). Is there any way to just send one 128k/s > > mountpoint and have the relay side transcode to the bitrates and go > > from there, that way I'm saving 80k/s of upload from the source > > host. > > The stream randomly starts to glitch out every once and awhile and > > I'm trying to fix it. Any pointers would be greatly appreciated. > > sure, sending the 128k stream works as normal and just have 2 local > transcoders to take the 128k stream from the relay, reencode and push > back to the relay under a different mountpoint. The likes of > streamtranscoder can do that, but I believe you can use a "curl | ices" > type solution as well. Not sure about darkice but it may do. > > karl > > >