I also sent mine in DE. I remember you also had the same problem with mce hardware error but you had reboots and all that. That was really a shame. It sounds like we had bad luck. I hope the next revision is better. I think i will wait. Did you send back your ram? The After-Sale Service said they are restocking and expecting it next week. Not sure though is that's already the new revision or not
Minisforum ms-01 i9-13900H 96GB with Proxmox
- Thread starter homelabenthusiast
- Start date
Notice: Page may contain affiliate links for which we may earn a small commission through services like Amazon Affiliates or Skimlinks.
I also sent mine in DE. I remember you also had the same problem with mce hardware error but you had reboots and all that. That was really a shame. It sounds like we had bad luck. I hope the next revision is better. I think i will wait. Did you send back your ram? The After-Sale Service said they are restocking and expecting it next week. Not sure though is that's already the new revision or not
nope, I sent back only the ms-01. I kept the RAM, SSD and the SFP+ transceiver. I’m really looking into buying a new one. But now wondering if should be worth wait for the new revision.
I bought an MS-01 to run Proxmox and I'm seeing constant hangs/reboots as well. I wonder if is something similar though I don't have the same exact errors you have. How did they determine it was a CPU issue?
I'm opting to return mine to Amazon if it is genuinely a hardware issue.
I'm opting to return mine to Amazon if it is genuinely a hardware issue.
I’ve attached all the error logs I got into Minisforum discord support forum and they told me it was a CPU issue. Since I bought trough Amazon DE they told me to ship it back and get the refund.
Maybe it’s something related to Proxmox? But other user are running it the same configuration without issues.
If you do not see the same error as mine, then it could be just proxmox or something else? Looking at the forum posts in proxmox i see a lot of people having the same problem. You can try to install other os like Linux mint or something... For my case, i sent them the logs and they told me its the CPU. I also bought a new RAM and saw the same problem... I also tried Linux Mint and i saw the same problem. So i sent it back
Microcode Install Instructions:
1) Install Proxmox 8.1
2) Add non-free-firmware to debian repo in sources.list
- Edit the /etc/apt/sources.list file. Add non-free-firmware to the 1st line so it looks like this---
- deb Index of /debian bookworm main contrib non-free-firmware
3) Save Changes
4) #apt clean && apt update
5) #apt install intel-microcode
- The current version Debian has in the repo is 3.2023114.1~deb12u1
6) Reboot, and the microcode patch should apply automatically.
7) You can check what microcode you are running after reboot by
grep 'stepping\|model\|microcode' /proc/cpuinfo
Running 2x MS-01’s with this patch, all good so far (all cores active)
Info/link to source:
1) Install Proxmox 8.1
2) Add non-free-firmware to debian repo in sources.list
- Edit the /etc/apt/sources.list file. Add non-free-firmware to the 1st line so it looks like this---
- deb Index of /debian bookworm main contrib non-free-firmware
3) Save Changes
4) #apt clean && apt update
5) #apt install intel-microcode
- The current version Debian has in the repo is 3.2023114.1~deb12u1
6) Reboot, and the microcode patch should apply automatically.
7) You can check what microcode you are running after reboot by
grep 'stepping\|model\|microcode' /proc/cpuinfo
Running 2x MS-01’s with this patch, all good so far (all cores active)
Info/link to source:
I'm now running 3 13900h nodes, with 96gb of RAM, 1xu.2 and 2x 22110 m. 2 drives. Running proxmox and ceph over a thunderbolt direct connection mesh. I haven't had any lockups. I did install the microcode patch as one of the first things I did.
I do think some of the issues others are seeing are due to bad hardware
I do think some of the issues others are seeing are due to bad hardware
Week average, the off time is when i moved one node. Light VM-duty as of now, 12 VM's, some peaks with Plex and Ollama, no iGPU passthrough at the moment.
Attachments
I opted to send my unit back to Amazon a few days before the return window. I had dealt with the issues for a few weeks only to see none of the changes not make any meaningful resolution. I actually watched the Craft Computing video before I bought the MS-01. Per the instructions, my microcode went from 0x410c to 0x411c. If this is not the right microcode, let me know what you guys have for your units. That microcode update didn't do anything for me, obviously. Loads aren't high at all with what I'm using it for (2 WinServer 2019 and 2 Win11)
Also reinstalled Proxmox a few times as I was essentially running a self administered crash course on it before I migrated my VMs.
I may try again with another MS-01 when the new batches are released. For now, I've installed Proxmox on my ancient NUC and so far I've not seen any unusual warnings.
Also reinstalled Proxmox a few times as I was essentially running a self administered crash course on it before I migrated my VMs.
I may try again with another MS-01 when the new batches are released. For now, I've installed Proxmox on my ancient NUC and so far I've not seen any unusual warnings.
Same here, returned while I still had time, although I didn't send back RAM, SFP+ transceiver and SSD.
But I really enjoyed it and the machine it's the real deal for homeserver enthusiasts. Currently I'm waiting for the next batch to be available in Amazon.
Nico
This sounds great. I'm interested in doing the same. Any pointers on setting up the thunderbolt direct connection mesh? I have not found much about this yet. Is it right that this limits the cluster to three nodes as you can't connect a fourth since there are only two USB4 ports per device?
Also would be interested in hearing what drives you went with; I was thinking of picking up an either a 3.84TB Samsung PM9A3 or a 4TB Intel U.2 on Ebay. Haven't looked into 22110 m.2s.
Just received my MS-01 added 96GB Memory and a NVMe M.2 drive, only AMI Bios v.2.22.0058 won't let me disable Secure Boot though it claims I'm at Access Level: Administrator, Secure Boot System Mode still claims: User and thus I guess it have grey out the option to disable Secure Boot, and thus it would let me install proxmox v.8, why the f..., any hints appreciated!
Attachments
Arh, after having installed a minimum Windoze 10 image, then I could disable Secure Boot and now boot off proxmox/fedora what have U
I also experienced hang/reboots running UNRAID 6.12.10 when the load exceeds the minimum. I ruled out all other possible issues and still can not find a fix.
I contacted after-sales support, and I was told to turn off the e-cores. I didn't deactivate those in the bios but pinned everything to the p-cores in UNRAID; now, the machine has become stable.
Right before posting, I also checked the microcode; it is 0x441c.
I guess I'm one of the unlucky ones, and this is a hardware problem.
I contacted after-sales support, and I was told to turn off the e-cores. I didn't deactivate those in the bios but pinned everything to the p-cores in UNRAID; now, the machine has become stable.
Right before posting, I also checked the microcode; it is 0x441c.
I guess I'm one of the unlucky ones, and this is a hardware problem.
You need enter bios, Before you try to enter any OS. This is new BIOS security standard, You cannot change any sucurity setting, after you try to enter OS.
Hi, If anyone meet "Memory failure: unhandlable page. / RAS" error.
It maybe faulty CPU. Please contact to your after sales support for return or change.
If problem could temporary fix with Turn off all small cores in BIOS setting. You can use this method to verify this problem.
It maybe faulty CPU. Please contact to your after sales support for return or change.
If problem could temporary fix with Turn off all small cores in BIOS setting. You can use this method to verify this problem.