ifconfig reports the MTU at 1500. It's a T1000, so yes - it's devouring one whole CPU/hardware-thread 100%.
Regards, Kevin On 10/12/06, Craig Bender <[EMAIL PROTECTED]> wrote:
It can be an issue if the MTU for the network is not 1500. Can you test it out? Is this a T2000? If so your Xsun process is taking up one whole CPU. Kevin Burtch wrote: > Only one user on the system, both processes below belong to him: > PID USERNAME SIZE RSS STATE PRI NICE TIME CPU PROCESS/NLWP > 14955 root 33M 31M cpu17 10 0 0:02:42 4.0% Xsun/1 > 15298 utcu0 11M 9560K sleep 59 0 0:00:05 0.1% uttsc-bin/2 > > utcapture output: > # TERMINALID TIMESTAMP TOTAL PACKET TOTAL LOSS BYTES SENT > PERCENT LOSS LATENCY > 00144f4892dc 20061012175713 9199 0 12729082 > > We just upgraded to a beta version of the Sun Ray 2 firmware (that > fixes other issues) in a shot-in-the-dark attempt to see if it would > help (it didn't, no change). > > I don't believe the MTU should be an issue, as the Sun Ray Server and > DTUs are all on the same private VLAN and I believe on the same > switch. > > Thanks, > Kevin > > > On 10/11/06, Craig Bender <[EMAIL PROTECTED]> wrote: >> Kevin, >> Does utcapture show anything. Have you tested the MTU. >> >> While this is happening, what does Xsun (runs as root) for this display >> show and what does uttsc-bin show? Can you get me some prstat -u UID >> and prstat -u root info (try to correlate the display for me to the Xsun >> process). >> >> Is there anyway you can help us recreate this? >> >> Kevin Burtch wrote: >> > >> > Both front-end and back-end (Sun Ray VLAN and LAN interfaces) are set >> > to auto resulting in 100/full on both the Suns and the switch ports >> > they are connected to. >> > (they don't have gig-e switches at that site) >> > >> > Output of netstat -i shows zero errors. >> > >> > Thanks, >> > Kevin >> > >> > On 10/11/06, *Alejandro Soler* <[EMAIL PROTECTED] >> > <mailto:[EMAIL PROTECTED]>> wrote: >> > >> > *Look at the network switch configuration for sunray, and the type >> > of conecction (if its full duplex, or half duplex)* >> > >> > *Cheers * >> > >> > *On Wed, 11 Oct 2006 13:11:53 +0000, Ross Morrison wrote* >> > >> > >> > > Kevin, >> > > >> > > I think I may have seen something similar to what you describe. >> > We had three Sun Rays connected via a gigabit switch to a V440 >> > machine. When we used certain graphic intensive programs the Sun >> > Ray's would become unresponsive and in some cases they would lose >> > connection to the server entirely. We resolved the issue by >> > limiting the network port on the V440 to 100MB to match that of >> > the Sun Rays and this resolved the issue. >> > > >> > > May not be the same but something to check? >> > > >> > > Cheers >> > > Ross >> > > >> > > Kevin Burtch wrote: >> >> This is a tough one. >> >> > We have a deployment on an island, very remote, using some >> in-house >> >> > developed software managing sensitive data. >> >> > I mention this, as I have never personally *seen* the issue, >> and I >> >> > won't be able to. We cannot move the software and data here for >> >> > reasons I can't go into. >> >> > >> >> > The problem is: >> >> > When a user generates a report, cells are being redrawn on the >> >> screen >> >> > (that's all I can get out of them). When this happens, they >> lose >> >> all >> >> > responsiveness with the mouse and keyboard until the program >> >> finishes >> >> > the report. >> >> > >> >> > CPU load is minuscule (single user on a T1000) on the Sun Ray >> >> Server. >> >> > The software itself is running on a Windows Terminal Server. >> If run >> >> > directly, or remotely via a windows PC, it runs in 15 seconds >> >> with no >> >> > loss in interactiveness. >> >> > When run via a Sun Ray, it takes a minimum of 80 seconds and >> all >> >> > interactiveness is lost. >> >> > >> >> > What I've tried: >> >> > Sun Ray Connector for Windows 1.0 (released version) >> >> > Sun Ray Connector for Windows 1.1 (beta version) >> >> > rdesktop 1.4.1 >> >> > rdesktop 1.5 >> >> > Applying every patch I can find related to X. >> >> > >> >> > All have the same problem. >> >> > This actually surprised me, as the symptoms are exactly the >> same as >> >> > those I DID see fixed in Connector 1.1 (we had that problem at >> >> another >> >> > site). Only the trigger seems to be different (it's not >> >> scrolling as >> >> > far as I know, just rapid repaints of cells). >> >> > >> >> > Anyone have any ideas? >> >> > >> >> > Thanks, >> >> > Kevin >> >> > _______________________________________________ >> >> > SunRay-Users mailing list >> >> > [email protected] <mailto:[email protected]> >> >> > http://www.filibeto.org/mailman/listinfo/sunray-users >> >> > >> >> > >> > >> > > >> > > -- >> > > >> > >> > > >> > >> > >> > _______________________________________________ >> > SunRay-Users mailing list >> > [email protected] <mailto:[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 _______________________________________________ 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
