I used to suffer a lot from it… but ever since I did a clean install of Rocky 8.5 (then easyupgrade to 8.7) and running 2024+, I’ve not seen it…
I use two Flames at my work, Mac Studio in Toronto office and I also have a remote Linux Flame that is in Montreal.
I have no micro-freezing on the Mac one.
The Linux Flame is just a remote machine that has no monitor connected to it. It also did not have a video card in it. I had no micro-freezing issues at first.
IT had to instal AJA in this Flame recently as I needed to do a live client session via Clearview so we needed an SDI out of it to connect.
Now I’m experiencing constant micro-freezing. I’m not saying it is the AJA but that’s the only thing that changed.
Linux Flame is 2023.2 on Centos 7.
Please make sure to update to the latest 2022.3.2, 2023.3.2 and 2024.0.2 Service Packs which fix issues related to micro freeze. The updates are fully compatible with the previous update so it should be a walk in the park. No need to archive or perform back up and no DKU update required so you should be good in a minute!
Please report here succes!
Just upgraded to 2022.3.2 (linux) and after a day of testing… Micro Freeze seems to be gone. Sooo happy
From my experience since several software updates I haven’t had the issue again. Linux.
Sometimes there is a momentary nanosecond lag, but I can abide by that.
Thanks a lot @moxxen for validating the fix! Very happy to hear this!
Unfortunately, we’re still getting it and I’m talking maybe 20 seconds or so.
2023.3.2.
and you are sure they are not S+W operations/autosaves?
Pretty sure
I don’t even like muttering the words “micro freeze” – since from my experience I thought this had been zapped or at least mitigated. But I’m getting them again, not like 20 seconds, just maybe 2-3 seconds. It happens about once a minute or so, and it is not during S+W operations, saves, etc.
2024.0; Linux threadripper.
Anyone else seeing this?
*This is in a new project, not much media in it. All media is locally cached.
Had them gone away in 2023, but they came back in only one project in 2024 (I asume it’s our framestore causing it atm as the project drains every frame from it and the upgrade to 2024.1 didn’t help). But it’s strange to have it happen again when all of a sudden at other places it is a thing again, too. Let’s see whether it is coincidence or sth else.
Mine are still 20 - 30 seconds.
I’m still on 2023. They seem to have gone away with the update 2023.3.2. There is a similar update for 2024. It does not need a new DKU, so it is a simple install.
I do have anew issue however, with 2023.3.2. About 1 in 3 times I use the motion vectors, I freeze. Not micro: more like permanent. I doesn’t actually crash, so I haven’t been able to generate a report. I haven’t had the time to wait it out. I’m going to try to hang it up when I leave tonight and see if it finally crashes.
Hmm. I haven’t seen that. But I’d stick with the thing that only messes up during one thing, over the thing that messes up everywhere.
Linux? Does the app freeze or the whole machine GUI?
What Flame version, and what GPU?
When you get this freeze, login via SSH, and type dmesg
It is very likely your GPU has crashed, look for nvidia errors.
Motion vectors have done this for me since they were introduced. I always save everything before adding a Motion Vectors Map.
Rocky Linux. Just the app freezes, but it can be slow and pokey to get back to the desktop. I generally close the terminal to crash it out. Last night before I left I tried to freeze it up and of course, I tried a hundred times and it worked perfectly. I had restarted the app just before I tried, so it might be some sort of unreleased memory issue. I will follow your advice next time it happens. For the record, I am not overstressing the system. The rez of the pieces I’m doing M-vectors on is generally about 600x600. Never had the problem until I installed 2023.3.2 (the microfreeze fix)
“Never had the problem until I installed 2023.3.2 (the microfreeze fix)”
Oh the irony!