baumix

Only setting Power Supply Idle Control wasn't enough. After almost 6 days it crashed again - I additionally disabled Global C-state Control (Advanced\AMD Overclocking\CPU Common Options\). I also uninstalled the UnraidConnect Plugin as mentioned in another thread for being a potential culprit, Uptime ist now 8 days 6 hours ... I hope the crashes are gone ...

Thanks for your reply, found the setting in BIOS (AdvancedßAMD CBS\CPU Common Options - Power Supply Idle Control: Default Low Current Idle) and set it accordingly (Typical Current Idle), might be reset by latest BIOS Update. We will see in a week or two. Wish me luck!

I have exactly one log from the crash last week - that led me to possible USB issues. But it was the only log I found ... on the other crashes, the log just ends. unraid-log_20250530.log.gz

Hi there, my Unraid server was running fine the last years using Docker containers, Windows and Linux VMs. A couple of weeks ago, I wanted to test some network settings because of some errors I discovered (they were WIndows related) and reactivated macvlan - never had issues in the past, but the known problems forced me to disable it a couple of months ago. Then problems started to arise - Unraid crashed and since then, the crashes arise once a week. I used a backup without macvlan, but the crashes didn't stop. I don't know if it's related to my experiments or if there are other issues I don't know about. I'm using Unraid 7.01 on an Asrock Taichi X370 (BIOS 10.41) pcie_acs_override=downstream, Windows 10/11 with GPU passthrough (GTX 1660 Super). Symptoms are: All VMs go down, Docker Containers are running fine, WebUI has three states: is not reachable prompts to set the initial root password states the flash drive has problems and needs to be replaced) ssh access is not working anymore. Since I have no access in this situation (GPU is passed through to Windows VM) and only have a black screen, the diagnostics cannot be retrieved - only afterwards. What have I done to resolve the issue: restored flash backup (network settings without using macvlan) formatted USB-Stick (without error messages) and restored flash backup changed USB Port for the flash drive The last crashes could be repaired by a (cold) restart ... I have no clue, where the issues are coming from, maybe someone can help me to identify the cause ... diagnostics enclosed. Cheers Baumix darwinstation-diagnostics-20250609-1948.zip

Of course! Just tossing an Asus X97-a with 2 PCI slots ... The spec reads Expansion Slots • 1 x PCIe 4.0 x16 (@ x16/x8/x4 mode) SafeSlot Core+ • 4 x PCIe 3.0 x16 (@ x1 mode) So the latter have a mechanical x16 slot, but electrically only x1 connected. It's all about the PCIe lanes a CPU provides ... A Ryzen 7950 for example provides 24 usable lanes ... they are used for M.2 Slots, SATA ports etc. expansion slots. And they even share lanes between e.g. M2 and PCIe, so you can only use either one. If you really need multiple x16 slots, a server board with a suitable CPU will be needed. They mostly have plenty of lanes. But maybe you tell us, which needs you have for multiple x16 slots ... Cheers Baumix

Everytime you restart the docker, it will check for game updates ...

I'm using the telegraf Docker and it's great. Will it be possible to include the ntpq binary, so ntp-stats could be collected? Cheers Baumix

Can you please share the changes you made? I may try a setup with Asus Prime X570-Pro, so how are your experiences with this board?