At 02:53 PM 30/10/2002 -0700, you wrote:>Bastard is used in a negative context here :) A snippit of a top >command reveals: > >PID USER PRI NI SIZE RSS SHARE STAT %CPU %MEM TIME COMMAND >3246 root 15 0 2436 2436 1280 S 10.5 3.9 0:14 ices > > >I am re encoding using: > > <encode> > <nominal-bitrate>8000</nominal-bitrate> > <samplerate>11025</samplerate> > <channels>1</channels> > <quality>-1</quality> > </encode> > >Seems like its chewing up a lot a cpu cycles or is this by design? >Considering you're doing on-the-fly downsampling and re-encoding, this is pretty good. ices is designed to be very efficient if all it's doing is streaming the content, but it's slow (for obvious reasons) if it has to decode and re-encode the data that it streams. Even worse, you're streaming in managed mode (ABR), which is less than half the speed of the quality-rated VBR modes. Michael <p>--- >8 ---- List archives: http://www.xiph.org/archives/ icecast project homepage: http://www.icecast.org/ To unsubscribe from this list, send a message to 'icecast-request@xiph.org' containing only the word 'unsubscribe' in the body. No subject is needed. Unsubscribe messages sent to the list will be ignored/filtered.
I've noticed with the cvs ices that the more memory which is leaked the higher the load average gets until just before all the ices processes get killed off by the kernel. During normal operations with three reencoding streams the average pretty much stays around .050 to 1.2. Kirk -- Kirk Reiser The Computer Braille Facility e-mail: kirk@braille.uwo.ca University of Western Ontario phone: (519) 661-3061 --- >8 ---- List archives: http://www.xiph.org/archives/ icecast project homepage: http://www.icecast.org/ To unsubscribe from this list, send a message to 'icecast-request@xiph.org' containing only the word 'unsubscribe' in the body. No subject is needed. Unsubscribe messages sent to the list will be ignored/filtered.
At 06:32 AM 31/10/2002 -0500, you wrote:>I've noticed with the cvs ices that the more memory which is leaked >the higher the load average gets until just before all the ices >processes get killed off by the kernel. During normal operations with >three reencoding streams the average pretty much stays around .050 to >1.2. >I've had no reports of memory leaks or unreasonably-high cpu usage. Perhaps you want to give a complete, detailed, bug report? (bugs.xiph.org) Michael <p>--- >8 ---- List archives: http://www.xiph.org/archives/ icecast project homepage: http://www.icecast.org/ To unsubscribe from this list, send a message to 'icecast-request@xiph.org' containing only the word 'unsubscribe' in the body. No subject is needed. Unsubscribe messages sent to the list will be ignored/filtered.