Closed kevmtan closed 1 day ago
Experiencing the same issue on Fedora Workstation 40 with Wayland, but not when running X11
Could you run in your terminal this command for a few seconds, paste the output into a file and upload it here?
nvidia-smi -q -x -lms 1000
thank you!
Not OP but facing same issue, here is my log:
Could you run in your terminal this command for a few seconds, paste the output into a file and upload it here?
nvidia-smi -q -x -lms 1000
thank you!
below is as much command output as I could fit within character limit... hope this helps!
many thanks for working on this :)
```
kt@dmn:~$ nvidia-smi -q -x -lms 1000
I also experience the problem on 24.04 with the Nvidia 560 drivers. Found this in my log:
meta_window_set_stack_position_no_sync: assertion 'window->stack_position >= 0' failed
Error
error@file:///home/gebruiker/.local/share/gnome-shell/extensions/monitor@astraext.github.io/src/utils/utils.js:147:23
updateNvidiaGpu@file:///home/gebruiker/.local/share/gnome-shell/extensions/monitor@astraext.github.io/src/gpu/gpuMonitor.js:1348:19
readOutput/</<@file:///home/gebruiker/.local/share/gnome-shell/extensions/monitor@astraext.github.io/src/utils/continuosTaskManager.js:89:37
readOutput/<@file:///home/gebruiker/.local/share/gnome-shell/extensions/monitor@astraext.github.io/src/utils/continuosTaskManager.js:88:40
@resource:///org/gnome/shell/ui/init.js:21:20
ERROR: updateNvidiaGpu: xml is undefined
edit: above happens when I select None as a gpu. When I select a Nvidia gpu I get some other error:
Unhandled promise rejection. To suppress this warning, add an error handler to your promise chain with .catch() or a try-catch block around your await expression. Stack trace of the failed promise:
_promisify/proto[asyncFunc]@resource:///org/gnome/gjs/modules/core/overrides/Gio.js:448:16
openExtensionPrefs@resource:///org/gnome/shell/ui/extensionSystem.js:321:26
openPreferences@resource:///org/gnome/shell/extensions/extension.js:27:26
addUtilityButtons/<@file:///home/gebruiker/.local/share/gnome-shell/extensions/monitor@astraext.github.io/src/menu.js:150:33
@resource:///org/gnome/shell/ui/init.js:21:20
This bug should be fixed in the new version that's gonna be released soon. Please share some feedback if you have any other issue regarding this or if everything is working fine.
@ljuzig I would like to test the v27 beta, but I'm not sure how to install the extension from a zip file. Extracting it to ~/.local/share/gnome-shell/extensions/monitor@astraext.github.io/
doesn't seem to work and gives errors. I'm currently running on Gnome 47. Plug-in version v26 does work when installed from the gnome extensions page.
Are there extra steps necessary to install the plug-in from a zip?
gnome-extensions install monitor@astraext.github.io.shell-extension.zip
should be enough.
It's in review anyway, if everything's going according to plan it should automatically update by tomorrow.
Thanks, that did the trick for installing. I'm still not getting any statistics for the nvidia gpu, only gpu information:
A dump of my nvidia-smi -q -x -lms 1000
:
nvidia-smi.txt
@veldenb your data works fine on my end:
You must have some sort of other error. Could you try enable debug on Astra Monitor settings, logout, log back in, open the GPU menu and close it 2/3 times, wait a couple of seconds and then share ~/.cache/astra-monitor/debug.log
? This would help a lot!
Thank you!
Nothing special in the Debug.log:
13:31:01.489 - Starting Processor Monitor monitoring
13:31:01.535 - Starting Gpu Monitor monitoring
13:31:01.546 - Starting Memory Monitor monitoring
13:31:01.574 - Starting Storage Monitor monitoring
13:31:01.575 - Starting Network Monitor monitoring
13:31:01.576 - AstraMonitor enabled
13:31:01.577 - Initializing container
13:31:01.577 - Creating Processor Header
13:31:01.582 - Creating Processor Menu
13:31:01.661 - Adding widget: processor
13:31:01.661 - Creating Processor Header
13:31:01.664 - Creating Gpu Menu
13:31:01.706 - Adding widget: gpu
13:31:01.706 - Creating Memory Header
13:31:01.708 - Creating Memory Menu
13:31:01.721 - Adding widget: memory
13:31:01.722 - Creating Storage Header
13:31:01.724 - Creating Storage Menu
13:31:01.724 - Initializing storage menu
13:31:01.734 - Adding widget: storage
13:31:01.734 - Creating Network Header
13:31:01.736 - Creating Network Menu
13:31:01.769 - Adding widget: network
13:31:01.769 - Creating Sensors Header
13:31:01.770 - Creating Sensors Menu
13:31:01.772 - Adding widget: sensors
13:31:01.772 - Creating Compact Header
13:31:01.772 - Adding widget: compact
13:31:03.006 - Updating Processor Monitor
13:31:03.042 - Updating Gpu Monitor
13:31:03.075 - Updating Network Monitor
13:31:03.773 - Placing container in right box at position 0
13:31:04.507 - Updating Processor Monitor
13:31:04.542 - Updating Gpu Monitor
13:31:04.546 - Updating Memory Monitor
13:31:04.575 - Updating Storage Monitor
13:31:04.592 - Updating Network Monitor
13:31:06.007 - Updating Processor Monitor
13:31:06.042 - Updating Gpu Monitor
13:31:06.092 - Updating Network Monitor
13:31:07.507 - Updating Processor Monitor
13:31:07.543 - Updating Gpu Monitor
13:31:07.546 - Updating Memory Monitor
13:31:07.575 - Updating Storage Monitor
13:31:07.592 - Updating Network Monitor
13:31:09.007 - Updating Processor Monitor
13:31:09.043 - Updating Gpu Monitor
13:31:09.092 - Updating Network Monitor
13:31:10.507 - Updating Processor Monitor
13:31:10.543 - Updating Gpu Monitor
13:31:10.546 - Updating Memory Monitor
13:31:10.575 - Updating Storage Monitor
13:31:10.592 - Updating Network Monitor
13:31:12.007 - Updating Processor Monitor
13:31:12.043 - Updating Gpu Monitor
13:31:12.092 - Updating Network Monitor
13:31:13.508 - Updating Processor Monitor
13:31:13.543 - Updating Gpu Monitor
13:31:13.546 - Updating Memory Monitor
13:31:13.575 - Updating Storage Monitor
13:31:13.595 - Updating Network Monitor
13:31:15.040 - Updating Processor Monitor
13:31:15.051 - Updating Gpu Monitor
13:31:15.112 - Updating Network Monitor
13:31:16.540 - Updating Processor Monitor
13:31:16.550 - Updating Memory Monitor
13:31:16.559 - Updating Gpu Monitor
13:31:16.578 - Updating Storage Monitor
13:31:16.613 - Updating Network Monitor
13:31:18.040 - Updating Processor Monitor
13:31:18.059 - Updating Gpu Monitor
13:31:18.113 - Updating Network Monitor
13:31:19.550 - Updating Processor Monitor
13:31:19.551 - Updating Memory Monitor
13:31:19.568 - Updating Gpu Monitor
13:31:19.591 - Updating Storage Monitor
13:31:19.616 - Updating Network Monitor
13:31:21.052 - Updating Processor Monitor
13:31:21.077 - Updating Gpu Monitor
13:31:21.120 - Updating Network Monitor
13:31:22.551 - Updating Memory Monitor
13:31:22.552 - Updating Processor Monitor
13:31:22.577 - Updating Gpu Monitor
13:31:22.591 - Updating Storage Monitor
13:31:22.620 - Updating Network Monitor
13:31:24.052 - Updating Processor Monitor
13:31:24.077 - Updating Gpu Monitor
13:31:24.120 - Updating Network Monitor
13:31:25.551 - Updating Memory Monitor
13:31:25.553 - Updating Processor Monitor
13:31:25.578 - Updating Gpu Monitor
13:31:25.591 - Updating Storage Monitor
13:31:25.620 - Updating Network Monitor
13:31:27.053 - Updating Processor Monitor
13:31:27.078 - Updating Gpu Monitor
13:31:27.120 - Updating Network Monitor
13:31:28.557 - Updating Processor Monitor
13:31:28.557 - Updating Memory Monitor
13:31:28.580 - Updating Gpu Monitor
13:31:28.599 - Updating Storage Monitor
13:31:28.642 - Updating Network Monitor
13:31:30.057 - Updating Processor Monitor
13:31:30.081 - Updating Gpu Monitor
13:31:30.142 - Updating Network Monitor
13:31:31.557 - Updating Processor Monitor
13:31:31.557 - Updating Memory Monitor
13:31:31.581 - Updating Gpu Monitor
13:31:31.599 - Updating Storage Monitor
13:31:31.642 - Updating Network Monitor
13:31:33.057 - Updating Processor Monitor
13:31:33.081 - Updating Gpu Monitor
13:31:33.143 - Updating Network Monitor
13:31:34.557 - Updating Memory Monitor
13:31:34.558 - Updating Processor Monitor
13:31:34.581 - Updating Gpu Monitor
13:31:34.599 - Updating Storage Monitor
13:31:34.643 - Updating Network Monitor
13:31:36.058 - Updating Processor Monitor
13:31:36.088 - Updating Gpu Monitor
13:31:36.143 - Updating Network Monitor
13:31:37.557 - Updating Memory Monitor
13:31:37.561 - Updating Processor Monitor
13:31:37.588 - Updating Gpu Monitor
13:31:37.599 - Updating Storage Monitor
13:31:37.643 - Updating Network Monitor
13:31:39.061 - Updating Processor Monitor
13:31:39.089 - Updating Gpu Monitor
13:31:39.143 - Updating Network Monitor
13:31:40.558 - Updating Memory Monitor
13:31:40.568 - Updating Processor Monitor
13:31:40.590 - Updating Gpu Monitor
13:31:40.599 - Updating Storage Monitor
13:31:40.643 - Updating Network Monitor
13:31:42.068 - Updating Processor Monitor
13:31:42.090 - Updating Gpu Monitor
13:31:42.143 - Updating Network Monitor
13:31:43.560 - Updating Memory Monitor
13:31:43.569 - Updating Processor Monitor
13:31:43.590 - Updating Gpu Monitor
13:31:43.603 - Updating Storage Monitor
13:31:43.643 - Updating Network Monitor
I do see some strange messages in journalctl:
sep 26 15:37:34 desktop kernel: __vm_enough_memory: pid: 47495, comm: nvidia-smi, bytes: 51539607552 not enough memory for the allocation
sep 26 15:37:34 desktop kernel: __vm_enough_memory: pid: 47495, comm: nvidia-smi, bytes: 51539644416 not enough memory for the allocation
sep 26 15:37:34 desktop kernel: __vm_enough_memory: pid: 47495, comm: nvidia-smi, bytes: 51539742720 not enough memory for the allocation
sep 26 15:37:37 desktop kernel: __vm_enough_memory: pid: 47495, comm: nvidia-smi, bytes: 51539607552 not enough memory for the allocation
sep 26 15:37:37 desktop kernel: __vm_enough_memory: pid: 47495, comm: nvidia-smi, bytes: 51539644416 not enough memory for the allocation
sep 26 15:37:37 desktop kernel: __vm_enough_memory: pid: 47495, comm: nvidia-smi, bytes: 51539742720 not enough memory for the allocation
sep 26 15:37:41 desktop kernel: __vm_enough_memory: pid: 47495, comm: nvidia-smi, bytes: 51539607552 not enough memory for the allocation
Timestamps are a bit off due to my local time zone (GMT+2:00) I think.
I don't think it's the nvidia-smi
command itself. I created a fake-nvidia-smi
command and added it to my $PATH
:
#!/bin/bash
while :
do
cat /tmp/nvidia-smi.txt
sleep 1
done
And then replaced the command in gpuMonitor.js
:
this.updateNvidiaGpuTask.start(`${path}fake-nvidia-smi -q -x -lms ${timer}`, {
flush: { trigger: '</nvidia_smi_log>' },
});
After logging in again I can then see it running in the background:
$ ps x | grep nvidia
109579 ? S 0:00 /bin/bash /home/veldenb/.local/bin/fake-nvidia-smi -q -x -lms 1400
But still same behaviour, no metrics with dummy input.
sep 26 15:37:34 desktop kernel: __vm_enough_memory: pid: 47495, comm: nvidia-smi, bytes: 51539607552 not enough memory for the allocation sep 26 15:37:34 desktop kernel: __vm_enough_memory: pid: 47495, comm: nvidia-smi, bytes: 51539644416 not enough memory for the allocation sep 26 15:37:34 desktop kernel: __vm_enough_memory: pid: 47495, comm: nvidia-smi, bytes: 51539742720 not enough memory for the allocation sep 26 15:37:37 desktop kernel: __vm_enough_memory: pid: 47495, comm: nvidia-smi, bytes: 51539607552 not enough memory for the allocation sep 26 15:37:37 desktop kernel: __vm_enough_memory: pid: 47495, comm: nvidia-smi, bytes: 51539644416 not enough memory for the allocation sep 26 15:37:37 desktop kernel: __vm_enough_memory: pid: 47495, comm: nvidia-smi, bytes: 51539742720 not enough memory for the allocation sep 26 15:37:41 desktop kernel: __vm_enough_memory: pid: 47495, comm: nvidia-smi, bytes: 51539607552 not enough memory for the allocation
This looks like a bug in nvidia-smi:
I'll make you a debug build with more information when I can just to make sure there is not something wrong in astra-monitor as well.
That would be great, I added some dump statements to the js files by hand. The nvidia xml is parsed correctly and added to the object tree. I think all works well except selecting the right gpu for the stats. When I select a gpu in the settings window the gpu gets selected but also kills the settings window. So maybe something goes wrong there when writing or reading the selected gpu.
I found a fix that works for me and created a PR for it: https://github.com/AstraExt/astra-monitor/pull/149 The bug will only show up if your PCI-id has a letter in it.
I'm going to issue a new release, thank you!
Description
GPU stats are completely blank for me, whether standalone graph, tooltip, or submenu within CPU. It seems that GPU stats cannot be retrieved by Astra at all, although GPU hardware info & addresses do show up. Other utilities like Nvidia Settings and GreenWithEnvy are working perfectly and show full stats. GTop is successfully detected & added to data sources list.
I've experienced this bug since I first installed Astra Monitor upon release of Ubuntu 24.04... Don't have any experience with prior versions.
Steps to Reproduce
Screenshots
Environment