Hi Craig, these are my Sun Ray Server settings you might be interested. I have also created a ticket, so support can digg into this.
-bash-3.00# tail /etc/system * optimize clock set hires_ticks=1 -bash-3.00# getopt CLK_TCK (after reboot) -- -bash-3.00# utxconfig -l -a DIMENSIONS=auto ENABLE_PSEUDOCOLOR_VISUAL=no DEFAULT_PSEUDOCOLOR_VISUAL=no ENABLE_MULTISCREEN=yes ENABLE_XINERAMA=no ENABLE_XKB=no SCREEN_GEOMETRY=auto SCREEN_ORDER=auto XSERVER_PREF=Xnewt QUICK_ABORT=yes XRENDER=yes regards, thomas -----Ursprüngliche Nachricht----- Von: [email protected] [mailto:[email protected]] Im Auftrag von Craig Bender Gesendet: Dienstag, 17. Mai 2011 15:16 An: SunRay-Users mailing list Betreff: Re: [SunRay-Users] 5.2 released ... slow display refresh Greetings Thomas, If you're not seeing packet loss, the hires change won't do anything. But to be sure it's in effect, what does getconf CLK_TCK report? Regarding the slow draw in general: Once again a request to open calls if you haven't done so and are seeing this issue. If you have opened one, ¡muchas gracias! This issue has been replicated by engineering and is being worked. Regarding the problem being client related, while that sounds logical given your tests, it doesn't rule out the server side. The interaction between the client, the server, and even the OS that is being displayed (due to things like Flash, RCA acceleration) all has to be looked at. All of it it percolates up through or somehow interacts with Xserver. In this case, I would agree that the issue appears below (above?) the RDP or X11 layer, meaning it happens regardless of what you are displaying, but the real problem could still be at the Xserver layer. On 5/17/11 5:49 AM, Fuerle, Thomas wrote: > ad slow display refresh: > > I have set /etc/system like 'set hires_tick=1', did not change much, at least > nothing I noticed. > > But somebody hinted running 5.2 with the old 5.1.2 firmware seems to > be a good approach. So I connected to my production Sun Ray server, > which is 5.1.2 and did a force of a downgrade of the firmware via > > utfwadm -A -e MAC -F -f /opt/SUNWut/lib/firmware_gui > > terminalID=MAC > terminalIPA=IP Adrress > model=SunRayP9 > currentAuth=current Auth server > currentFW=GUI4.2_140993-07_2010.12.30.17.25 > currentBarrier=422 > currentBarrierLevel=422 > currentMTU=1500 > Subnet=255.255.0.0 > Router=current Router > LeaseTim=604800 > DHCPServer=current DHCP > INFORMServer=current INFORM > tftpSrvr=current tftpSrvr > FWservType=conf > speed=100F > parmsVersion=GUI4.2_140993-07_2010.12.30.17.25 > parmsBarrier=422 > configMTU=1500 > dnsList=current DNS servers > dname=current dname > confNetType=DHCP > confTftpSrvr=current tftp > conf.fulldup=1 > confServers=current conf server > stopqon=0 > bandwidth=100000000 > poweroff=30 > > did go back to the firmware GUI and changed server to servername of srs5.2, > while keept firmware server to the current prod 5.1.2 and rebootet. > > so far this combination seems to work nice as I already know it, which > is subjective, but that's just my feeling, > > So my conclusion would be that something in the firmware is the hand brake or > broken, which would explain, why it is independent from the client OS you > running to connect to. > > regards, thomas > > -bash-3.00# > -----Ursprüngliche Nachricht----- > Von: [email protected] > [mailto:[email protected]] Im Auftrag von Craig Bender > Gesendet: Montag, 16. Mai 2011 17:20 > An: [email protected] > Betreff: Re: [SunRay-Users] 5.2 released > > Got it. Sorry, I incorrectly assumed Putty was only available for > Windows. Is it lame that gnome-terminal meets my needs? ;) > > First, if you have support, please open a case so we can track it from a > development side. > > There are a few people responding, all with different scenarios, which all > could be different issues. For Win7/2008R2 via MS-RDP, it's possible that > RCA is kicking in. Adding -D to uttsc would be an interesting test to run. > > I believe there are some new features in 5.2 that could affect drawing, > though I would expect them to be positive impacts. I want to double check > with the developer before I claim new features that may or not have made the > cut as I don't seem them in the release notes, nor do I see a way to change > the one that is suspect in my mind. > > In the meantime, if folks could gather the following information. If you are > not comfortable with putting this information on the list, please give the > information to Oracle Support. > > -Note if settings /etc/system like 'set hires_tick=1' are enabled. > -Any settings in the parms files (if used) -Provide a utquery of DTUs > exhibiting this behavior -Describe the network, both on server and > client side -Any packet loss detected with utcapture -MTU settings > agreed upon between server and client...and whether that matches a > ping test for max MTU > (http://blogs.oracle.com/ThinkThin/entry/the_importance_of_mtu) > > On 5/16/11 6:53 AM, Philippe Bourdeu d'Aguerre wrote: >> Le lundi 16 mai 2011, [email protected] a écrit : >>> Isn't putty windows? >>> >> I have a plain (not virtalized) CentOS server running SRSS 5.2 and >> Sunrays connecting to it with Gnome or KDE CentOS sessions on the same >> server. >> Refresh is randomly slow. > _______________________________________________ > SunRay-Users mailing list > [email protected] > http://www.filibeto.org/mailman/listinfo/sunray-users > _______________________________________________ > SunRay-Users mailing list > [email protected] > http://www.filibeto.org/mailman/listinfo/sunray-users _______________________________________________ SunRay-Users mailing list [email protected] http://www.filibeto.org/mailman/listinfo/sunray-users _______________________________________________ SunRay-Users mailing list [email protected] http://www.filibeto.org/mailman/listinfo/sunray-users
