This seems to have just happened with a recent update, not sure. As of about a few days ago.
Just generally using the interface, any program, the computer stops responding for like half a second or a full second, then resumes. Annoyingly, it doesn’t buffer the mouse clicks, so when clicking around files and folders, it’s annoying having to wait and then nothing happens, and then trying again.
It does buffer the keyboard input though, so when I’m typing in the Kate text editor, if nothing happens for a bit, I wait and the line of text I typed will appear.
I would say this happens fairly frequently, like, every 10 seconds, though at random.
Music continues to play uninterrupted in the background.
Does this sound like any known issue?
Is there any profiling software I can use that would show a lag spike, perhaps in CPU usage, or maybe even HDD access?
Happened to me one time two days before after my reinstallation of eos with Plasma. I have no idea what this could be due to. I went to the terminal via ctrl and F1 or F2 and rebooted. Never happened again since then
Well I can see you got some dated hardware there on the HDD side of things.
Especially the ID-4: /dev/sdc vendor: Seagate model: ST2000DL003-9VT166 size: 1.82 TiBdrive that seems to have been available since end 2010 (not saying you bought it back than, but still).
So it might be a good idea to do have a look at the smart status of the drives.
Well I don’t know if there is any important data on that drive , but I would back it up as soon as possible. The early signs may point to some flaws on the drive, or maybe even some bit rot.
Running a deeper smart control test might more or less destroy your data, so as I said before you do that, back up your data.
Yep that’s why I bought it.
It’s not those kind of lags, it’s a full interface lockup, feels quite buggy.
Eg. if it were a performance bottleneck, particularly the HDD, I would expect the file explorer to hitch when I opened a folder, but not while typing text in to Kate. This lock up affects every program, the entire interface.
Thanks, I didn’t know about that one, I was trying htop and couldn’t see anything obvious, like, Kate itself would jump to the top. Which isn’t helpful. But I would get the same lockup in, say, Discord or anything else.
BUT… today the problem isn’t there. So I’ll keep an eye on it.
Thing is, the lock up is like 1 or 2 seconds at a time. Not sure top refreshes fast enough to catch that momentary spike. What I’d need is something that shows a performance graph and I could click on the spike and see the program name that caused the spike. Do we have anything like that?
You could also try System Monitor, which does have graphing, but as far as I can tell, won’t let you hone in on a spike to get details. It may yet provide an idea of what resource is being maxed though. Eg: GPU, memory, CPU, etc.
So I have taken a look at the Bios date of your Mobo and found that though it is from 2018 that would be the latest available. So no update would be available , but I also saw that the first Bios available was from 2015. So if I am correct the system would be from around 2015. Do you know if the PSU is from around that time also, and is a solid brand. Another thing is your PC connected directly to a (preferably grounded) wall socket , if not try to change that. The reason I am saying this is because you mentioned the 200% spike could be caused by some power peak that isn’t handled correctly by your PSU because of it’s age or something like that. To monitor your system there is something called glances that runs in the terminal and gives you a lot of info.
It’s a fairly recently replaced PSU from memory, I just checked and it’s an Antec EDGE 550 (Model: EDG550) I know it’s recent enough, there’s no way that’s an issue.
Thanks for glance, it still updates on intervals like top so no way I’d catch the 1-2 second spike without a graph I’d say.
I thought one of the best things about Linux is that it is good for exactly that…