From: Roland Schulz (roland_at_utk.edu)
Date: Thu Sep 23 2010 - 23:13:03 CDT

Hi,

I haven't tested it myself but according to the people from LRZ(1) VMD works
good over VirtualGL (2).
For non-3D applications I have the best experience with NX (3-5). But I'm
not sure how well this works for 3D.

1) http://www.lrz.de/services/compute/visualisation/visualisation_8/
2) http://www.virtualgl.org/
3) http://www.nomachine.com/download.php
4) http://freenx.berlios.de/
5) code.google.com/p/neatx

Roland

On Tue, Sep 14, 2010 at 12:03 PM, Moulton, Steve <moultonsa_at_ornl.gov> wrote:

> Hi John,
>
> What I really wanted was a silver bullet. No such luck.
>
> At this point, we are going to ask users to migrate their data from this
> system when they need to run VMD. If we run into a roadblock, I'll check
> back about other possible solutions.
>
> - Steve
>
> -----Original Message-----
> From: John Stone [mailto:johns_at_ks.uiuc.edu]
> Sent: Monday, September 13, 2010 5:35 PM
> To: Moulton, Steve
> Cc: 'vmd-l_at_ks.uiuc.edu'
> Subject: Re: vmd-l: GLXBadLargeRequest and BadLength messages
>
> Hi,
> I can't comment on the X/GLX OpenGL remote display issues other
> than to say that this has been an issue that crops up somewhat frequently
> under different versions of linux as bugs have crept into the X11 server
> implementations at various times. I've found this particularly frustrating
> as most of the classic unix systems from SGI, Sun, and others never had
> these
> kinds of problems. I suspect that X/GLX OpenGL remote display is simply
> one of the features that the maintainers of modern linux distributions
> don't test very carefully when they update to a new X11 version.
>
> Regarding potential workarounds:
> In most cases you'll get much better performance running VMD
> locally than via a network connection. Can you load files from
> your 10.2 system over a network filesystem of some kind?
> The molecular structure and trajectory files usually contain
> less data than the resulting stream of OpenGL commands and geometry
> that get generated from them, so you would get both better display
> performance and use far less network bandwidth to transfer your
> data to your display machine rather than remote displaying.
>
> If there's some other hardware or security limitation that causes
> you to want to remote display, let me know and I may be able to
> suggest other alternative strategies.
>
> Cheers,
> John Stone
> vmd_at_ks.uiuc.edu
>
> On Mon, Sep 13, 2010 at 04:43:32PM -0400, Moulton, Steve wrote:
> > Greetings,
> >
> >
> >
> > While running VMD 1.8.7 on a OpenSuSE 10.2 64 bit system but
> displaying on
> > an OpenSuSE 11.2 64 bit system, I am seeing many messages of the form
> >
> >
> >
> > XRequest.135: GLXBadLargeRequest 0x2
> >
> > XRequest.135: BadLength (poly request too large or internal Xlib
> length
> > error) 0x2
> >
> >
> >
> > For various reasons, it is not possible to update the 10.2 system.
> VMD
> > ran correctly when the display system was OpenSuSE 11.0, but these
> systems
> > are being upgraded. Setting VMDSIMPLEGRAPHICS is a workaround, but
> is
> > not quite satisfactory.
> >
> >
> >
> > Anyone have any ideas how to get around this one? I did try throwing
> a
> > VNC to no avail, and the X servers don't have the LBX extension (which
> use
> > to be quite effective at patching up X protocol errors).
> >
> >
> >
> > Thanks,
> >
> >
> >
> > Steve Moulton
> >
> > moultonsa_at_ornl.gov
>
> --
> NIH Resource for Macromolecular Modeling and Bioinformatics
> Beckman Institute for Advanced Science and Technology
> University of Illinois, 405 N. Mathews Ave, Urbana, IL 61801
> Email: johns_at_ks.uiuc.edu Phone: 217-244-3349
> WWW: http://www.ks.uiuc.edu/~johns/ Fax: 217-244-6078
>
>

-- 
ORNL/UT Center for Molecular Biophysics cmb.ornl.gov
865-241-1537, ORNL PO BOX 2008 MS6309