sywor

I have tried this and it does not work for me. It keeps resetting back to 45/55 C on the default setting page as well.

I'm trying to update my server from 6.11.0-rc3 to 6.11.0-rc5 but it keeps failing with: "run failed: /bin/bash" Very cryptic failure. I have retried several times, with the same result every time. tower-diagnostics-20220930-1819.zip

Just booted into version 6.9.0-beta35. Lets see if it helps.

The fault is back, this time the machine stayed up for 10 days. And the syslog is full of something suspicious I cannot decipher. It starts at 02:07:13 this morning (Nov 21) on row 358. Is my CPU broken? syslog

No that's not what I wrote. I cannot export diagnostics when it happens since it is by then it is already too late to do anything. Anyways. I have now turned on syslog persistence to flash. hoppfull at least the syslog should be saved when it happens again.

Bumping my own thread. The problem is still there even in safe mode. And since the kworker consumes all the CPU the diagnostics never completes, even when I run it from the console.

Booted into safe mode, started the array and let it run for an hour. So far everything seams normal as far as I can see. tower-diagnostics-20201031-1353.zip

I had to reboot the machine because now even the console where unresponsive. After reboot I manage to collect the diagnostics. tower-diagnostics-20201031-1233.zip Edit: Note that I did not start the array when the collected this diagnostics file. It just occurred to me that it might be required. Should I collect another diagnostics file or is the one above enough?

Hi! Recently my unraid machine have started become unresponsive after approximately 3 days after reboot. As you can see in the attached image the kworker events_power_efficient consumes all the cpu and there are several other kworker kblocked processes filling up the rest of the top consuming processes. The web GUI is unresponsive so I don't know how to save the System Diagnostics. Worth noting is that the machine has been running non-stop perfectly for four months before this started happening earlier this week.

I think it is working now. The machine has been stable for a few hours. I went through the BIOS settings one more time and deeep in there I think I found the correct c-state settings.

Bump! I still need help with this! 🥺

I have not been able to find any settings in the BIOS that mentions anything with "Power Supply Idle Control" or "C State" The only thing I could find related to sleep / hibernate was this: And it was already set to off. When booting into 6.8.3 it is still unstable and it spits out the following trace log:

I already have the latest BIOS I will try to change the BIOS settings next time I reboot. I also tried to download 6.6.0, 6.7.0 but the url trick you wrote about does not work. it would appear the then only host the last two versions that are listed on the donwload site. ex: https://s3.amazonaws.com/dnld.lime-technology.com/stable/unRAIDServer-6.6.0-x86_64.zip is dead

The only changes I have made to the BIOS is enable virtualization. I'm not running any VMs that has explicit pass-through for any hardware. All I could find on unraid and ryzen bios tweaks are almost one year old and concerns pass-through of graphic cards etc. Are there any other bios tweaks you would recomend?

I reverted back to 6.5.3 last night and the machine has been stable all night. But I rely want to upgrade to the latest version since my new nic does not show up in the old version and 6.5.3 is rely old at this point

I have been stuck on version 6.5.3 for ages since the motherboard I used had a Marvel controller that is unsupported in later versions (I think there should be a few threads about it on this forum). So I recently upgrade my system with completely new hardware (except the discs). I made sure to read the forum to find a HBA that others got working, LSI 9305-i24. Worth noting is that the system is stable on 6.5.3 with the new hardware, it ran for 14+ hours and even completed a parity check. With 6.8.3 is is not staying up for more then a few minutes. I have included a diagnostics file that I just barely managed to download before the system crashed again, The hardware XML, and a few pictures for what the screen looked like after it crashed. tower-diagnostics-20200418-1138.zip unraid_hw.xml