Script for automated testing of computer latency.
14 Mar 2024 07:24 - 14 Mar 2024 07:28 #295879
by seuchato
Replied by seuchato on topic Script for automated testing of computer latency.
Klopp
Thanks for the flowers
1) correct, and very well answered explained by cornholio, too
2) See this article . Imho it explains jitter and standard deviation well. I'second your conclusion. But that's also depending on use case:
Soft stepping: From my understanding, own experience and reading, I'd conclude, that about 50.000 uS is the community agreed limit. I myself use this PC for my small router on parallel port BOB. It tested around 20.000 uS. In my daily use, I gave it a safety margin, so I actually run it using 50.000 uS for Jitter.
Hard stepping: There are numerous opinions on a reasonable limit. I guess everything under 100.000 can be considered good. tommylight made some statements, that he could see even 200.000 as usable. Kindly note: I lack experience on this, so that's only my finding from reading in LCNC forums.
3) Jitter is the figure to watch. I'd second your conclusion, but be cautious and set some safety margin. Citing tommylight again: "use LinuxCNC for several hours. If no warnings = nice "
4) see above: about 100.000 should be fine
greez
chris
Thanks for the flowers
1) correct, and very well answered explained by cornholio, too
2) See this article . Imho it explains jitter and standard deviation well. I'second your conclusion. But that's also depending on use case:
Soft stepping: From my understanding, own experience and reading, I'd conclude, that about 50.000 uS is the community agreed limit. I myself use this PC for my small router on parallel port BOB. It tested around 20.000 uS. In my daily use, I gave it a safety margin, so I actually run it using 50.000 uS for Jitter.
Hard stepping: There are numerous opinions on a reasonable limit. I guess everything under 100.000 can be considered good. tommylight made some statements, that he could see even 200.000 as usable. Kindly note: I lack experience on this, so that's only my finding from reading in LCNC forums.
3) Jitter is the figure to watch. I'd second your conclusion, but be cautious and set some safety margin. Citing tommylight again: "use LinuxCNC for several hours. If no warnings = nice "
4) see above: about 100.000 should be fine
greez
chris
Last edit: 14 Mar 2024 07:28 by seuchato. Reason: Timeout
Please Log in or Create an account to join the conversation.
- jg00163206
- Offline
- Premium Member
Less
More
- Posts: 92
- Thank you received: 4
22 May 2024 05:29 #301129
by jg00163206
Replied by jg00163206 on topic Script for automated testing of computer latency.
Using user defined parameters:
Result: running as user: mill,
runtime: 7200 seconds,
parameters: --nobase, --base 100000,
prefix for outfiles 1716355190
nglx: 5,
LinuxCNC pc tests, version 20200930, started 22.05.2024 01:21:08
********************************************************************************
Hardware info is logged here: 1716355190.hw.
Latency data is logged here: 1716355190.lat.
General info:
Dell Inc., Vostro 270
Bios version A08, dated 11/23/2012
Chipset: Intel Corporation 7 Series/C216
Harddisk is rotational
The amount of swap currently used is 0
================================================================================
CPU related data:
model name : Intel(R) Core(TM) i5-3570 CPU @ 3.40GHz
cpu cores : 4
stepping : 9
cache size : 6144 KB
================================================================================
RAM related data:
Maximum Capacity: 8 GB
Size: 8 GB
Size: 8 GB
================================================================================
GPU related data:
product: Oland GL [FirePro W2100]
product: Xeon E3-1200 v2/3rd Gen Core processor Graphics Controller
configuration: depth=32 driver=radeon latency=0 resolution=1920,1080
configuration: driver=i915 latency=0
================================================================================
Parallel port related data:
number of parallel ports is 0
================================================================================
Os and desktop related data:
PRETTY_NAME="Debian GNU/Linux 12 (bookworm)"
Should the info below not match with your machine,
kindly post the output of pstree here, including a brief
desciption of your desktop environment, window manager and
display manager. Tia
forum.linuxcnc.org/18-computer/39370-scr...-of-computer-latency
Desktop environment : lightdm
Windowmanager : Xfwm4
Displaymanager : lightdm
================================================================================
Kernel related data:
Kernel 6.1.0-18-rt-amd64
GRUB_CMDLINE_LINUX_DEFAULT="quiet"
Cpu idle driver: intel_idle
================================================================================
Kernel modules data:
Check if pcspkr is loaded: ./lcnc-hw: line 576: [: too many arguments
Yes, may cause bigger latency.
can be eliminated using:
sudo echo "install pcspkr /bin/true" >/etc/modprobe.d/pcspkr.conf
Check if snd modules are loaded: Yes, may cause bigger latency.
can probably be disabled in bios, alternatively, you can run
"./block-snd-modules"
which has just now been created for your convenience.
================================================================================
Keyboard & Mouse related data:
Number of mice: 1
Mouse 0 : Telink Wireless Receiver Mouse
Mice attached to USB:
./lcnc-hw: line 631: lsusb: command not found
Keyboards attached to USB:
./lcnc-hw: line 633: lsusb: command not found
================================================================================
LinuxCNC related data:
LinuxCNC version is: 2.9.2
================================================================================
Preparing for latency test:
Started glxgears No. 1, 2, 3, 4, 5
......ATTENTION: default value of option mesa_glthread overridden by environment.
ATTENTION: default value of option mesa_glthread overridden by environment.
ATTENTION: default value of option mesa_glthread overridden by environment.
....
Started firefox with 'vimeo.com/150574260', WID
================================================================================
STartime = Wed May 22 01:21:22 AM EDT 2024
Command for latency test is:
su mill -c "latency-histogram --nobase --base 100000 --sbinsize 1000 --nox 2>&1 | tee -a 1716355190.lat &"
Latency testing loop started Wed May 22 01:21:22 AM EDT 2024,
should end after Wed May 22 03:21:22 AM EDT 2024
Core 0: +49.0°C (high = +85.0°C, crit = +105.0°C)
Core 1: +48.0°C (high = +85.0°C, crit = +105.0°C)
Core 2: +46.0°C (high = +85.0°C, crit = +105.0°C)
Core 3: +46.0°C (high = +85.0°C, crit = +105.0°C)
done command, retval = 0
.Note: Using POSIX realtime
0 s servo min: -4.400 us max: 5.400 us sdev: 1.600 us
. 1 s servo min: -4.400 us max: 7.100 us sdev: 1.600 us
2 s servo min: -5.500 us max: 7.600 us sdev: 1.600 us
. 2 s servo min: -7.700 us max: 10.100 us sdev: 1.700 us
. 3 s servo min: -7.700 us max: 10.100 us sdev: 1.700 us
4 s servo min: -7.700 us max: 10.100 us sdev: 1.600 us
. 4 s servo min: -7.700 us max: 10.100 us sdev: 1.600 us
5 s servo min: -7.700 us max: 10.100 us sdev: 1.700 us
There are no windows in the stack
Invalid window '%1'
Usage: windowactivate [options] [window=%1]
--sync - only exit once the window is active (is visible + active)
If no window is given, %1 is used. See WINDOW STACK in xdotool(1)
Remaining 0d 1h:59m:55s, press ^C to abort...
6 s servo min: -15.200 us max: 13.200 us sdev: 1.700 us
6 s servo min: -15.200 us max: 13.200 us sdev: 1.700 us
7 s servo min: -15.200 us max: 13.200 us sdev: 1.700 us
7 s servo min: -15.200 us max: 13.200 us sdev: 1.700 us
8 s servo min: -15.200 us max: 13.200 us sdev: 1.700 us
9 s servo min: -23.900 us max: 24.300 us sdev: 1.800 us
9 s servo min: -23.900 us max: 24.300 us sdev: 1.800 us
Core 0: +57.0°C (high = +85.0°C, crit = +105.0°C)
Core 1: +57.0°C (high = +85.0°C, crit = +105.0°C)
Core 2: +54.0°C (high = +85.0°C, crit = +105.0°C)
Core 3: +53.0°C (high = +85.0°C, crit = +105.0°C)
Remaining 0d 1h:59m:50s, press ^C to abort...
10 s servo min: -23.900 us max: 24.300 us sdev: 1.800 us
11 s servo min: -23.900 us max: 24.300 us sdev: 1.800 us
11 s servo min: -23.900 us max: 24.300 us sdev: 1.800 us
12 s servo min: -23.900 us max: 24.300 us sdev: 1.800 us
13 s servo min: -23.900 us max: 24.300 us sdev: 1.700 us
13 s servo min: -23.900 us max: 24.300 us sdev: 1.700 us
14 s servo min: -23.900 us max: 24.300 us sdev: 1.700 us
15 s servo min: -23.900 us max: 24.300 us sdev: 1.700 us
Core 0: +59.0°C (high = +85.0°C, crit = +105.0°C)
Core 1: +58.0°C (high = +85.0°C, crit = +105.0°C)
Core 2: +56.0°C (high = +85.0°C, crit = +105.0°C)
Core 3: +54.0°C (high = +85.0°C, crit = +105.0°C)
Remaining 0d 1h:59m:45s, press ^C to abort...
15 s servo min: -49.500 us max: 35.900 us sdev: 1.700 us
16 s servo min: -49.500 us max: 35.900 us sdev: 1.700 us
17 s servo min: -49.500 us max: 35.900 us sdev: 1.700 us
17 s servo min: -49.500 us max: 35.900 us sdev: 1.700 us
18 s servo min: -49.500 us max: 35.900 us sdev: 1.700 us
19 s servo min: -49.500 us max: 35.900 us sdev: 1.700 us
19 s servo min: -49.500 us max: 35.900 us sdev: 1.700 us
20 s servo min: -49.500 us max: 35.900 us sdev: 1.700 us
Core 0: +61.0°C (high = +85.0°C, crit = +105.0°C)
Core 1: +59.0°C (high = +85.0°C, crit = +105.0°C)
Core 2: +55.0°C (high = +85.0°C, crit = +105.0°C)
Core 3: +56.0°C (high = +85.0°C, crit = +105.0°C)
Remaining 0d 1h:59m:40s, press ^C to abort...
Result: running as user: mill,
runtime: 7200 seconds,
parameters: --nobase, --base 100000,
prefix for outfiles 1716355190
nglx: 5,
LinuxCNC pc tests, version 20200930, started 22.05.2024 01:21:08
********************************************************************************
Hardware info is logged here: 1716355190.hw.
Latency data is logged here: 1716355190.lat.
General info:
Dell Inc., Vostro 270
Bios version A08, dated 11/23/2012
Chipset: Intel Corporation 7 Series/C216
Harddisk is rotational
The amount of swap currently used is 0
================================================================================
CPU related data:
model name : Intel(R) Core(TM) i5-3570 CPU @ 3.40GHz
cpu cores : 4
stepping : 9
cache size : 6144 KB
================================================================================
RAM related data:
Maximum Capacity: 8 GB
Size: 8 GB
Size: 8 GB
================================================================================
GPU related data:
product: Oland GL [FirePro W2100]
product: Xeon E3-1200 v2/3rd Gen Core processor Graphics Controller
configuration: depth=32 driver=radeon latency=0 resolution=1920,1080
configuration: driver=i915 latency=0
================================================================================
Parallel port related data:
number of parallel ports is 0
================================================================================
Os and desktop related data:
PRETTY_NAME="Debian GNU/Linux 12 (bookworm)"
Should the info below not match with your machine,
kindly post the output of pstree here, including a brief
desciption of your desktop environment, window manager and
display manager. Tia
forum.linuxcnc.org/18-computer/39370-scr...-of-computer-latency
Desktop environment : lightdm
Windowmanager : Xfwm4
Displaymanager : lightdm
================================================================================
Kernel related data:
Kernel 6.1.0-18-rt-amd64
GRUB_CMDLINE_LINUX_DEFAULT="quiet"
Cpu idle driver: intel_idle
================================================================================
Kernel modules data:
Check if pcspkr is loaded: ./lcnc-hw: line 576: [: too many arguments
Yes, may cause bigger latency.
can be eliminated using:
sudo echo "install pcspkr /bin/true" >/etc/modprobe.d/pcspkr.conf
Check if snd modules are loaded: Yes, may cause bigger latency.
can probably be disabled in bios, alternatively, you can run
"./block-snd-modules"
which has just now been created for your convenience.
================================================================================
Keyboard & Mouse related data:
Number of mice: 1
Mouse 0 : Telink Wireless Receiver Mouse
Mice attached to USB:
./lcnc-hw: line 631: lsusb: command not found
Keyboards attached to USB:
./lcnc-hw: line 633: lsusb: command not found
================================================================================
LinuxCNC related data:
LinuxCNC version is: 2.9.2
================================================================================
Preparing for latency test:
Started glxgears No. 1, 2, 3, 4, 5
......ATTENTION: default value of option mesa_glthread overridden by environment.
ATTENTION: default value of option mesa_glthread overridden by environment.
ATTENTION: default value of option mesa_glthread overridden by environment.
....
Started firefox with 'vimeo.com/150574260', WID
================================================================================
STartime = Wed May 22 01:21:22 AM EDT 2024
Command for latency test is:
su mill -c "latency-histogram --nobase --base 100000 --sbinsize 1000 --nox 2>&1 | tee -a 1716355190.lat &"
Latency testing loop started Wed May 22 01:21:22 AM EDT 2024,
should end after Wed May 22 03:21:22 AM EDT 2024
Core 0: +49.0°C (high = +85.0°C, crit = +105.0°C)
Core 1: +48.0°C (high = +85.0°C, crit = +105.0°C)
Core 2: +46.0°C (high = +85.0°C, crit = +105.0°C)
Core 3: +46.0°C (high = +85.0°C, crit = +105.0°C)
done command, retval = 0
.Note: Using POSIX realtime
0 s servo min: -4.400 us max: 5.400 us sdev: 1.600 us
. 1 s servo min: -4.400 us max: 7.100 us sdev: 1.600 us
2 s servo min: -5.500 us max: 7.600 us sdev: 1.600 us
. 2 s servo min: -7.700 us max: 10.100 us sdev: 1.700 us
. 3 s servo min: -7.700 us max: 10.100 us sdev: 1.700 us
4 s servo min: -7.700 us max: 10.100 us sdev: 1.600 us
. 4 s servo min: -7.700 us max: 10.100 us sdev: 1.600 us
5 s servo min: -7.700 us max: 10.100 us sdev: 1.700 us
There are no windows in the stack
Invalid window '%1'
Usage: windowactivate [options] [window=%1]
--sync - only exit once the window is active (is visible + active)
If no window is given, %1 is used. See WINDOW STACK in xdotool(1)
Remaining 0d 1h:59m:55s, press ^C to abort...
6 s servo min: -15.200 us max: 13.200 us sdev: 1.700 us
6 s servo min: -15.200 us max: 13.200 us sdev: 1.700 us
7 s servo min: -15.200 us max: 13.200 us sdev: 1.700 us
7 s servo min: -15.200 us max: 13.200 us sdev: 1.700 us
8 s servo min: -15.200 us max: 13.200 us sdev: 1.700 us
9 s servo min: -23.900 us max: 24.300 us sdev: 1.800 us
9 s servo min: -23.900 us max: 24.300 us sdev: 1.800 us
Core 0: +57.0°C (high = +85.0°C, crit = +105.0°C)
Core 1: +57.0°C (high = +85.0°C, crit = +105.0°C)
Core 2: +54.0°C (high = +85.0°C, crit = +105.0°C)
Core 3: +53.0°C (high = +85.0°C, crit = +105.0°C)
Remaining 0d 1h:59m:50s, press ^C to abort...
10 s servo min: -23.900 us max: 24.300 us sdev: 1.800 us
11 s servo min: -23.900 us max: 24.300 us sdev: 1.800 us
11 s servo min: -23.900 us max: 24.300 us sdev: 1.800 us
12 s servo min: -23.900 us max: 24.300 us sdev: 1.800 us
13 s servo min: -23.900 us max: 24.300 us sdev: 1.700 us
13 s servo min: -23.900 us max: 24.300 us sdev: 1.700 us
14 s servo min: -23.900 us max: 24.300 us sdev: 1.700 us
15 s servo min: -23.900 us max: 24.300 us sdev: 1.700 us
Core 0: +59.0°C (high = +85.0°C, crit = +105.0°C)
Core 1: +58.0°C (high = +85.0°C, crit = +105.0°C)
Core 2: +56.0°C (high = +85.0°C, crit = +105.0°C)
Core 3: +54.0°C (high = +85.0°C, crit = +105.0°C)
Remaining 0d 1h:59m:45s, press ^C to abort...
15 s servo min: -49.500 us max: 35.900 us sdev: 1.700 us
16 s servo min: -49.500 us max: 35.900 us sdev: 1.700 us
17 s servo min: -49.500 us max: 35.900 us sdev: 1.700 us
17 s servo min: -49.500 us max: 35.900 us sdev: 1.700 us
18 s servo min: -49.500 us max: 35.900 us sdev: 1.700 us
19 s servo min: -49.500 us max: 35.900 us sdev: 1.700 us
19 s servo min: -49.500 us max: 35.900 us sdev: 1.700 us
20 s servo min: -49.500 us max: 35.900 us sdev: 1.700 us
Core 0: +61.0°C (high = +85.0°C, crit = +105.0°C)
Core 1: +59.0°C (high = +85.0°C, crit = +105.0°C)
Core 2: +55.0°C (high = +85.0°C, crit = +105.0°C)
Core 3: +56.0°C (high = +85.0°C, crit = +105.0°C)
Remaining 0d 1h:59m:40s, press ^C to abort...
Please Log in or Create an account to join the conversation.
22 May 2024 19:04 #301159
by seuchato
Replied by seuchato on topic Script for automated testing of computer latency.
jg00163206
Did you press ^C to abort?
If, a bit a pitty, as it does not look off too far, given you have still screws to turn to optimise.
Greez
chris
Did you press ^C to abort?
If, a bit a pitty, as it does not look off too far, given you have still screws to turn to optimise.
Greez
chris
Please Log in or Create an account to join the conversation.
Time to create page: 0.080 seconds