On 03/03/10 22:31, DRC wrote: > ti...@piments.com wrote: >> > VNC context: >> > >> > All tiger vnc software is run on the remote kubuntu box via ssh from >> > local linux system. >> > >> > r...@local# ssh -C -X -L 5900:localhost:5900 remote.dyndns.info >> > r...@remote:~# vncviewer localhost:0 >> >> >> I'm wondering whether running this way means the vnc compression is >> irrelevant to the network section since it is happening between two >> processes on the same machine. > > That isn't really the way it works. The VNC compression is still > occurring between the local and remote machines. It's just that the > compressed VNC stream is being re-processed (encrypted and > re-compressed) by the SSh tunnel. > > However, I'm not sure if TigerVNC behaves this way, but I know that > other VNC implementations will detect if a connection is to localhost > and automatically default to using Raw compression (that is, no > compression.)
Thanks for the reply. I'm using command line arguments to turn off auto detection and set sensible values since auto does not work well in this context. Just for the record auto was not setting raw but 24b colour with jpeg 6 compression. > That seems to be the way your configuration is behaving, > based on the fact that SSh compression seems to speed things up (which > it shouldn't do if your image stream was actually being compressed by VNC.) > Well there's about an order of magnitude speed difference if I don't use -C !! That's what led me to think it was coming raw down ssh link. > One thing we never established is whether JPEG compression is active for > your VNC sessions. Whenever you drop the JPEG quality down to, say, 2 > or 3, do you see noticeable JPEG compression artifacts? If not, then > JPEG is not being used, and that's why it's running so slow. > yes, running ssh -C , jpeg quality can be seen to change and F8 connection information shows "requested encoding" changes according to what I set on command line or via options dlg. All compression options seem about equal in this context, but changing from 24b to 256 colours makes about a factor of 2 difference. In contrast , if I run a small X program directly from a ssh -C login terminal it is very slow to display the slightest change although the bandwidth it uses is minimal (2-3kB/s), I'm guessing there is a heavy latency doing it that way. Thanks again. > >> It does run with expected performance running viewer on client machine >> via an insecure port >> but that is unacceptable. >> >> Is there a better way? > > ------------------------------------------------------------------------------ Download Intel® Parallel Studio Eval Try the new software tools for yourself. Speed compiling, find bugs proactively, and fine-tune applications for parallel performance. See why Intel Parallel Studio got high marks during beta. http://p.sf.net/sfu/intel-sw-dev _______________________________________________ Tigervnc-devel mailing list Tigervnc-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/tigervnc-devel
