Gentle persons:
Recently, I wrote...
> I just ran 15-minute latency tests on my ASUS AT5NM10-I board (which is
> equipped with an Intel Atom D510 cpu) first with a stock boot and then with
> isolcpus=1 as a boot parameter. I've posted my results in the table on the
> Wiki.
>
> In brief, isolating one cpu shaved approximately 6000ns off the max jitters
> reported for both the base thread and the servo thread.
>
> Of course the latency test should be run much longer but I believe these
> numbers are indicative. YMMV
I had a few minutes this afternoon and decided to try another experiment with
this board---still running Ubuntu 10.04LTS with the 2.6.32-122-rtai kernel.
I disabled Gnome (which shut down X) on the board by running the following from
the command line:
sudo service gdm stop
(I disabled a few other services too, like Apache, but I don't think this is
relevant. The stock Ubuntu/EMC2 install starts up far too many services for my
taste. When I want a controller, that's all I want, not a desktop computer.)
Then I logged out and disconnected keyboard/mouse/monitor (well, actually I
just set my KVM switch to another computer; possibly it still imposes some
electrical levels at the motherboard connectors.)
Finally, I logged into the board via Ethernet from another computer using ssh
-Y and again ran the latency-test for 15 minutes with approximately the same
level of stress as before (2 copies of glxgear, web browsing, listing the
directories of an external USB drive, etc.).
Bottom line: the latency numbers got even better. Max jitter fell to
3211ns/3222ns.
For some time I have envisioned running my tabletop mill with a headless
controller and a networked operator console. If anything, the present result
says this would be the preferred mode. Of course, I could make it cleaner by
stripping down the distribution some more, possible even install RTAI and EMC2
over a Ubuntu server distribution to avoid the Gnome/X-server stuff altogether,
since there's the usual niff-naff about booting Ubuntu desktop edition without
a monitor attached, etc.
I'm curious if anyone has tried this experiment with other boards and, if so,
what results were obtained.
And, yes, I realize I am chasing after diminishing returns. The latency on a
stock board is already good enough for practical purposes. Let's just say my
home situation causes a form of ADHD were I work furiously on very short-term
projects.
Regards,
Kent
------------------------------------------------------------------------------
Ridiculously easy VDI. With Citrix VDI-in-a-Box, you don't need a complex
infrastructure or vast IT resources to deliver seamless, secure access to
virtual desktops. With this all-in-one solution, easily deploy virtual
desktops for less than the cost of PCs and save 60% on VDI infrastructure
costs. Try it free! http://p.sf.net/sfu/Citrix-VDIinabox
_______________________________________________
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users
