Sun Rays can't handle packet fragmentation - they just drop fragmented packets. Adjust the MTU down until you don't get any fragmentation through your ATM link.
The best option is to use a ping that supports packet size and the don't fragment option and keep dropping the packet size until it reliably gets through without fragmentation, then put the MTU option into your DHCP servers for the remote subnet. You shouldn't need to drop the MTU on the Sun Ray server. Paul On 11/11/05, Thomas Portmann <[EMAIL PROTECTED]> wrote: > Hello there, > > We have SRSS 3.0 on 2 Sparc/Solaris servers and have local Sun Rays as well > as remote Sun Rays, accessing the servers over an ATM link. The remote users > are complaining about poor performance: > 1. slow display updates > 2. sometimes a long delay between the input of characters and the display > in StarOffice of these characters. > > The WAN link (ATM, 40mbps) has been tested and is good. > > We run: > - Solaris 9 on Sparc (sun4u) > - WUNWuto is version 3.0_51 > - the usage of the link hardly exceeds 8mbps (sometimes up to 20mbps for a > very short period). > - the latency is usualy 10ms with single peaks at up to 80ms > > Do I need to say that localy we do not have these problems? > > The question is: is there something we overlooked and is causing this bad > user response? because we basically thought that having a low latency and a > big enough bandwidth would be sufficient. > > Thanks for any input/idea. > > > tom > > > PS: link usage and latency measured with Cacti/rrdtool (5 minute average) > > _______________________________________________ > 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
