System stops for several seconds

General system Newbie
1

I notice it happens the system stops for some seconds every now and then. Actually it seems not to be a full stopping most of the time, as I can usually move the mouse. It seems like working through commands is delayed or paused, as clicks are cached and done if the system reacts again.
I can’t tell how to reproduce, as it happened when I opened the Bash, while typing a comment in a forum (typed letters appeared after some seconds), opening a new Tab in FF, trying out gaming performance (game froze, sound moved on, I could move the mouse, then then 10 or so seconds happened, just like having a great, big lag), even just clicking on things on the welcome screen.

I’m using KDE, installed yesterday and checked for updates yesterday. I am very fond of EOS so far and liking it, but those “forced breaks” are a bit unnerving. Anything I can check or do, or any point what might let this happen and how to resolve it? I am thankful for any pointer or hint.

System Specs:
CPU: AMD 3700X,
GPU: AMD 5700 XT,
128GB OCZ Agility 3 SSD,
32GB RAM.

Thanks in advance

2

Well there could be a couple of things wrong. Off the top of my head there was the amd cpu stutter problem if your bios isn’t up to date (assuming it gets updates). I also remember a youtuber by the name of Brodie Robertson discovering a bug that causes lags when using any type of effect such as alt + tab, the tiler, etc. This occurs because the .cache directory is on a slow drive or when a drive is really busy.

See this bug here:
https://bugs.kde.org/show_bug.cgi?id=487043

Lastly, I would monitor resources at the time to see if anything looks unusual when these events occur.

3

I updated the BIOS, but it seems to still happen. I have a couple of those in the log:

[Di, 28. Mai 2024, 12:18:29] ata3: SATA link up 1.5 Gbps (SStatus 113 SControl 310)
[Di, 28. Mai 2024, 12:18:29] ata3.00: configured for UDMA/33
[Di, 28. Mai 2024, 12:18:55] ata3: SATA link up 1.5 Gbps (SStatus 113 SControl 310)
[Di, 28. Mai 2024, 12:18:55] ata3.00: configured for UDMA/33
[Di, 28. Mai 2024, 12:19:50] ata3.00: exception Emask 0x0 SAct 0x0 SErr 0xd0000 action 0x6 frozen
[Di, 28. Mai 2024, 12:19:50] ata3: SError: { PHYRdyChg CommWake 10B8B }
[Di, 28. Mai 2024, 12:19:50] ata3.00: failed command: WRITE DMA
[Di, 28. Mai 2024, 12:19:50] ata3.00: cmd ca/00:08:88:fc:9c/00:00:00:00:00/e3 tag 11 dma 4096 out
res 40/00:00:00:00:00/00:00:00:00:00/00 Emask 0x4 (timeout)
[Di, 28. Mai 2024, 12:19:50] ata3.00: status: { DRDY }
[Di, 28. Mai 2024, 12:19:50] ata3: hard resetting link
[Di, 28. Mai 2024, 12:19:50] ata3: SATA link up 1.5 Gbps (SStatus 113 SControl 310)
[Di, 28. Mai 2024, 12:19:51] ata3.00: configured for UDMA/33
[Di, 28. Mai 2024, 12:19:51] ata3.00: device reported invalid CHS sector 0
[Di, 28. Mai 2024, 12:19:51] ata3: EH complete
[Di, 28. Mai 2024, 12:19:59] ata3: SATA link up 1.5 Gbps (SStatus 113 SControl 310)
[Di, 28. Mai 2024, 12:19:59] ata3.00: configured for UDMA/33
[Di, 28. Mai 2024, 12:20:06] ata3: SATA link up 1.5 Gbps (SStatus 113 SControl 310)
[Di, 28. Mai 2024, 12:20:06] ata3.00: configured for UDMA/33
[Di, 28. Mai 2024, 12:20:14] ata3: SATA link up 1.5 Gbps (SStatus 113 SControl 310)
[Di, 28. Mai 2024, 12:20:14] ata3.00: configured for UDMA/33
[Di, 28. Mai 2024, 12:20:34] ata3: SATA link up 1.5 Gbps (SStatus 113 SControl 310)
[Di, 28. Mai 2024, 12:20:34] ata3.00: configured for UDMA/33
[Di, 28. Mai 2024, 12:20:51] ata3: SATA link up 1.5 Gbps (SStatus 113 SControl 310)
[Di, 28. Mai 2024, 12:20:51] ata3.00: configured for UDMA/33
[Di, 28. Mai 2024, 12:21:51] ata3.00: exception Emask 0x0 SAct 0x0 SErr 0xd0000 action 0x6 frozen
[Di, 28. Mai 2024, 12:21:51] ata3: SError: { PHYRdyChg CommWake 10B8B }
[Di, 28. Mai 2024, 12:21:51] ata3.00: failed command: WRITE DMA
[Di, 28. Mai 2024, 12:21:51] ata3.00: cmd ca/00:20:20:c3:2d/00:00:00:00:00/e0 tag 24 dma 16384 out
res 40/00:01:ea:bf:cd/00:00:1d:00:00/e0 Emask 0x4 (timeout)
[Di, 28. Mai 2024, 12:21:51] ata3.00: status: { DRDY }
[Di, 28. Mai 2024, 12:21:51] ata3: hard resetting link
[Di, 28. Mai 2024, 12:21:52] ata3: SATA link up 1.5 Gbps (SStatus 113 SControl 310)
[Di, 28. Mai 2024, 12:21:52] ata3.00: configured for UDMA/33
[Di, 28. Mai 2024, 12:21:52] ata3: EH complete

It seems there is a problem with the SATA?

4

Yeah, you could try unplugging ssd and plugging it back in or using a different SATA cable if it uses a cable. I’m not sure what to do about hardware problems.

5

Will try that. Also found something about the system trying to use multipath, but SATA doesn’t use it, so it times out and fails. But that one was for Ubuntu, so I can’t say if there is something similar as multipath.conf to remove the drive from multipath.

Edit: Switched ports at the mainboard to no avail.

6

Swap the cables too between the drives, the errors might appear for different drives after. I’ve googled few of the errors failed command: WRITE DMA and device reported invalid CHS sector 0 all point to drive or cable issues. If you have some cables lying around, you should try those too. I had a similar problem when one of the sata cables weren’t connected properly.

1 Like
7

32 GB RAM? But you do have swap? Description reminds me of my OOM situations, when not using swap.

From my understanding you just installed EoS for the first time? So either fresh system or windows earlier? This would rather exclude disk failure, which should be visible.

Otherwise as others pointed, test the drive

8

I sounds similar to the delay encountered when a hard disk drive spins up.

If you don’t have a hard disk drive in your system, might there be something in your system connected to a remote system (eg: SFTP), or network drive (NAS, SMB / NFS / SFTP, etc)?

9

Now I will find out the hard way if it was the old SSD or the cable or it is something within my system. :smiley:
Since I wanted to switch to Endeavour anyway I thought I spare myself the fiddling with cables and just resize the Windows partition and give Endeavour the main part. Downloaded a recommended application for the resize and it turns out it messed up and Windows didn’t boot anymore. A whole day of trying to restore the BCD and bring Windows up again it came to me Windows is a sore loser, reformated the (current) SSD and now Endeavour is officially the only system. So I will see if the hiccups still occur, or if it was the old SSD.

It is just my HDD doesn’t appear in Dolphin anymore. I really have to access it, as it has my backup data and mails… This would be my next find. Finding my backup data on the HDD and restore my mails to Linux Thunderbird and my savegames to Grim Dawn…

Ok, now I am really into EandeavourOS. It turns out EOS didn’t ignore a partition, but Windows Recovery just stupidly worked on Disk 1, erasing my data grave and backups, as my old Windows partition was on Disk 2. So had a not working skeleton of Windows written over my backup data and all my general data. Shivering in fear and being frustrated as hell I searched a bit and found a tutorial for testdisk. Well, testdisk couldn’t recover the partition of my data grave, but now is merrily copying my backup and general data. when it is finished I will wipe that old NTFS disk completely and restore my data grave and backups on a nice Linux drive making my PC totally MS free. Thanks EOS for saving years of data. now checks NAS prices for backups.

Now I am officially migrating to EOS!

10

I have found out that I got the same problem.
I encountered it on kernel >= 6.9.* but I do not have the issue on kernels <= 6.8.*.
I downgraded to 6.8 for now, hopefully it will be fixed on 6.10.

11

Some notes about system stopping for a few seconds. These might not be relevant to this particular case, but I guess worth mentioning what I’ve seen on my system.

  • If one or more disks are no more available (see fstab, udisks2, udiskie, etc.), then some program may think it is still available, and can cause such delay. This happened to me with Dolphin a few days ago.
  • If the default kernel is the reason, then maybe the LTS kernel works, worth a try.
12

If everything on the system runs fine on linux-lts, I would think that is a better choice.
6.8 kernel series is EOL and will no longer receive bugfix and security updates.
The LTS will.

3 Likes
13

There were no changes to the drives, so the kernel is the only cause of the problem.

14

It was the fastest way as I already had it in cache. The 6.10 should release soon enough and I will try that first before going to LTS and probably submitting a bug if it doesn’t yet exist.

15

It must have been something with either the SSD, the SATA connection (I could check that those points by mounting the drive. Didn’t have need for that so far) or the installation was missing something (I can’t remember I was asked for a swap file at the first trial installation. Perhaps that drive was too small). After my full install on my main SSD, with swap file I have no problems at all. All is running smoothly. No hiccups, no stops, from Libre Office over gimp to Heroic Games Launcher and Retroarch all is running well.
Right now I am feeling right at home with Linux and EOS. :slight_smile:

3 Likes