OS become laggy after some time of use having only integrated video

User Support
1

I started to use a new laptop and it has only integrated video Intel Corporation Metero Lake-P (Intel Arc Graphics)

After a reboot, it has some minutes when it is working good (not perfect at all, it is laggy from the beginning but nothing scary)
But after aroud 20-30 minutes of use (browsing, vscode … nothing resource consumtion at all) it starts to become more and more laggy

I’m having another laptop where I had this issue but it had a dedicated Nvidia video and after I blacklisted it everything went perfectly fine.

Here I don’t have one so not sure why I’m having the same issue

2

Try to use kernel-latest in dom0 (kernel version 6.11 currently).
Also check the dom0 logs to see if there are any errors using journalctl -b command.

3

I’m already on latest 6.11

Some messages that are yellow or red are:

  • CPU topo: CPU limit of 16 reached. Ignoring further CPUs (yellow)
  • ACPI BIOS wargning bus: incorrect checksum in table [bgrt] … (yellow)
  • [firmware bug]: cpu : APIC ID mismatch. CPUID: … (red, lots of them)
  • [firmware bug]: cpu : APIC ID mismatch. firmware: … (red, lots of them)
  • cpu 0 spinlock event irq (yellow, lots of them)
  • pnp number:number: disabling [mem …] because it overlaps … bar [yellow, lots of them]
4

Maybe it’s overheating?
Check the temperature sensors when it starts lagging in Q → Gear icon → Other → Sensor Viewer.

5

Every time I’m running sensors command into dom0 in terminal I see nothing that I should worry there

6

Yesterday when started to be extremely laggy it was when I started a presentation into the google browser
But as I said this type of issues I had only with dedicated graphic cards now I don’t have one anymore

7

Was it lagging in just this one qube or in all qubes or even in dom0?

8

I’ll came back with an answer after I’ll have more investigations
When it was extremely laggy didn’t interacted with dom0 nor other vms (other vms were opened but not interacted with them in that time)

9

Also check the memory usage in the affected qube using free -m command in its terminal to make sure that it’s not swapping.

10

Did you try to give the qube more RAM?

11

It is not because of that 99% sure.
It already has 16Gb of ram and 6 CPU

I’m using just browser, vscode [where just coding nothing fancy] and terminal

12

Regarding sensors I saw nvme at 60C (usually below 50 at @45) when became extramely laggy would that be the problem?
Regarding resources all the time I’m having more than enough

13

60C shouldn’t be a problem for a disk.
Do you have the same temperature rise by 10C for other sensors as well?
If it’s just nvme that is heating up then it could mean that there is a high IO load for nvme and this could be the reason for lagging.
Maybe try to monitor the disk IO.

14

That would make no sense as I read/write nothing on them when starts to lagg

More or less I saw just on nvme this big increase with +10,15C
The problem is that it starts to be so laggy that investigating in that period of time is almost impossible
(dom0 for example is working well)

15

All qubes have a swap space, if you over provision memory in qubes and one start to use a lot of memory, all the other may swap.

16

After some time while using it daily (work laptop) I founded the problem.

The browser, from time to time, starts to eat all of the cpu. Reason is teams (running in browser) , but why is doing that is not clear.

It is what it is, I need teams (work laptop) so …

Thanks

17

Try the unofficial heavy client, it works a lot better for me.

1 Like
18

I do expect to work better but it has lots of trackers so I prefer to stay with browser whenever I can

19

Is there something that I can investigate why is this happening?

So it is because of teams 100% sure, but starts to become laggy more and more whenever I’m on calls, other than that it works ok (teams is always opened for chat - running on browser not as a standalone app - browser is FF ESR 128.4)