Minis forum MS-A2

Notice: Page may contain affiliate links for which we may earn a small commission through services like Amazon Affiliates or Skimlinks.

johanqwerty

New Member
Mar 8, 2024
5
0
1
7945HX PBO -500 Ubuntu Server with 3 NVMe's (no heatsinks) populated:
mt7921_phy0-pci-0600
temp1: +40.0°C
nvme-pci-0700
Composite: +41.9°C
Sensor 1: +41.9°C
Sensor 2: +47.9°C
nvme-pci-0800
Composite: +51.9°C
Sensor 1: +51.9°C
Sensor 2: +38.9°C
amdgpu-pci-0100
vddgfx: 805.00 mV
vddnb: 844.00 mV
edge: +44.0°C
PPT: 16.12 W
k10temp-pci-00c3
Tctl: +48.6°C
Tccd1: +41.5°C
Tccd2: +41.4°C
nvme-pci-0900
Composite: +51.9°C
Sensor 1: +51.9°C
Sensor 2: +38.9°C
 
Last edited:

JaxJiang

Member
Jan 10, 2023
88
80
18
Is Silent Mode better then Performance Mode for the Fans?
I would assume that the silent mode doesnt cool better then performance.
Silent Mode.MAX CPU frequency lower than default.
Most of idle or low load high temperture, cause by very high max cpu frequency.
 

ike2117

New Member
Aug 4, 2025
6
0
1
Has anyone ran into an issue in proxmox where the entire system freezes? When checking the previous boot in journalctl -b -1 -xe was

Code:
░░ Subject: A start job for unit pve-daily-update.service has begun execution
░░ Defined-By: systemd
░░ Support: https://www.debian.org/support
░░
░░ A start job for unit pve-daily-update.service has begun execution.
░░
░░ The job identifier is 599.
Aug 04 05:52:06 pve pveupdate[381821]: <root@pam> starting task UPID:pve:0005D393:0060C9FA:689082C6:aptupdate::root@pam:
Aug 04 05:52:06 pve pveupdate[381843]: update new package list: /var/lib/pve-manager/pkgupdates
Aug 04 05:52:07 pve pveupdate[381821]: <root@pam> end task UPID:pve:0005D393:0060C9FA:689082C6:aptupdate::root@pam: OK
Aug 04 05:52:07 pve systemd[1]: pve-daily-update.service: Deactivated successfully.
░░ Subject: Unit succeeded
░░ Defined-By: systemd
░░ Support: https://www.debian.org/support
░░
░░ The unit pve-daily-update.service has successfully entered the 'dead' state.
Aug 04 05:52:07 pve systemd[1]: Finished pve-daily-update.service - Daily PVE download activities.
░░ Subject: A start job for unit pve-daily-update.service has finished successfully
░░ Defined-By: systemd
░░ Support: https://www.debian.org/support
░░
░░ A start job for unit pve-daily-update.service has finished successfully.
░░
░░ The job identifier is 599.
Aug 04 05:52:07 pve systemd[1]: pve-daily-update.service: Consumed 1.985s CPU time.
░░ Subject: Resources consumed by unit runtime
░░ Defined-By: systemd
░░ Support: https://www.debian.org/support
░░
░░ The unit pve-daily-update.service completed and consumed the indicated resources.

Afterwards, the entire system was stuck. Unable to access web interface, unable to ping, unable to ssh, tried checking output to a kvm but it wasn't on at the time so there was no display. A restart brought the device back, but still odd that it happened. Readings in the sensors seem fine, seemingly not dealing with heat issues. Latest BIOS, latest PVE. Some other posts online suggest may be hardware related, but I'm not entirely sold on the idea
 

VivienM

Member
Jul 7, 2024
36
8
8
Toronto, ON
Has anyone ran into an issue in proxmox where the entire system freezes? When checking the previous boot in journalctl -b -1 -xe was

Code:
░░ Subject: A start job for unit pve-daily-update.service has begun execution
░░ Defined-By: systemd
░░ Support: https://www.debian.org/support
░░
░░ A start job for unit pve-daily-update.service has begun execution.
░░
░░ The job identifier is 599.
Aug 04 05:52:06 pve pveupdate[381821]: <root@pam> starting task UPID:pve:0005D393:0060C9FA:689082C6:aptupdate::root@pam:
Aug 04 05:52:06 pve pveupdate[381843]: update new package list: /var/lib/pve-manager/pkgupdates
Aug 04 05:52:07 pve pveupdate[381821]: <root@pam> end task UPID:pve:0005D393:0060C9FA:689082C6:aptupdate::root@pam: OK
Aug 04 05:52:07 pve systemd[1]: pve-daily-update.service: Deactivated successfully.
░░ Subject: Unit succeeded
░░ Defined-By: systemd
░░ Support: https://www.debian.org/support
░░
░░ The unit pve-daily-update.service has successfully entered the 'dead' state.
Aug 04 05:52:07 pve systemd[1]: Finished pve-daily-update.service - Daily PVE download activities.
░░ Subject: A start job for unit pve-daily-update.service has finished successfully
░░ Defined-By: systemd
░░ Support: https://www.debian.org/support
░░
░░ A start job for unit pve-daily-update.service has finished successfully.
░░
░░ The job identifier is 599.
Aug 04 05:52:07 pve systemd[1]: pve-daily-update.service: Consumed 1.985s CPU time.
░░ Subject: Resources consumed by unit runtime
░░ Defined-By: systemd
░░ Support: https://www.debian.org/support
░░
░░ The unit pve-daily-update.service completed and consumed the indicated resources.

Afterwards, the entire system was stuck. Unable to access web interface, unable to ping, unable to ssh, tried checking output to a kvm but it wasn't on at the time so there was no display. A restart brought the device back, but still odd that it happened. Readings in the sensors seem fine, seemingly not dealing with heat issues. Latest BIOS, latest PVE. Some other posts online suggest may be hardware related, but I'm not entirely sold on the idea
Are you running the 6.8 or 6.14 kernel?
 

ike2117

New Member
Aug 4, 2025
6
0
1
I have no idea if it would help with your issue, but I'd suggest the 6.14. I went straight to the 6.14 on mine; my view is that the AMD 9xxx is new enough that you probably don't want a kernel from early 2024.
I have the 7945hx model but I'll still give 6.14 a try. There have been 2 of these freezes so far so hopefully this solve
 

MadFerIt

New Member
Jun 29, 2025
9
6
3
I have the 7945hx model but I'll still give 6.14 a try. There have been 2 of these freezes so far so hopefully this solve
7945hx here with the 128GB DDR5 Crucial kit here and a -750 pbo max boost reduction, no freezes so far. Kept the kernel / software up to date on the non-enterprise channel (from the very start), currently 6.8.12-13-pve as of yesterday. But my utilization is fairly low on the CPU besides a few openwebui/ollama tests.
 

ike2117

New Member
Aug 4, 2025
6
0
1
7945hx here with the 128GB DDR5 Crucial kit here and a -750 pbo max boost reduction, no freezes so far. Kept the kernel / software up to date on the non-enterprise channel (from the very start), currently 6.8.12-13-pve as of yesterday. But my utilization is fairly low on the CPU besides a few openwebui/ollama tests.
Same 128gb kit -500 pbo. This crash happened with all VMs shutdown. If my issue persists with the new kernel, I'll adjust pbo further down.
 

MadFerIt

New Member
Jun 29, 2025
9
6
3
Same 128gb kit -500 pbo. This crash happened with all VMs shutdown. If my issue persists with the new kernel, I'll adjust pbo further down.
Would definitely run a memtest / stress test to see if any errors occur, may need a replacement 128gb kit. I really need to get around to doing it myself too just in case I haven't triggered anything yet.
 

ike2117

New Member
Aug 4, 2025
6
0
1
Would definitely run a memtest / stress test to see if any errors occur, may need a replacement 128gb kit. I really need to get around to doing it myself too just in case I haven't triggered anything yet.
Any specific test you would recommend?
 

PnoT

Active Member
Mar 1, 2015
661
162
43
Texas
Anyone using the X710 NICs in a SET team in Windows Server? I'm using the latest drivers and firmware, and every time I team the adapters up, the switch fails to send network traffic if used by a VM. Immediately after creation, there is a brief window of network, and then nothing. I've run across multiple articles detailing issues with the LLDP agent being turned on in the BIOS, so I have disabled that, but am still having problems.

If I create a virtual switch with just 1 NIC, everything works normally, but what's the point in that when you need converged.

This seems to be directly related to my problem, but no solution: No SET-Switch Team possible on Intel X710 NICs? | Microsoft Community Hub

Intel says that in the 700 series, it's enabled in firmware by default, and in the MS-A2 BIOS, I can confirm it was set to enabled until I disabled it. Firmware Link Layer Discovery Protocol (FW-LLDP) - 30.3 - ID:705831 | Intel® Ethernet Adapters and Devices User Guide

This PDF states how to enable/disable it via different OS and methods: Intel® Ethernet Controllers FW LLDP Usage Control
 
Last edited:

tempest

New Member
Jun 18, 2025
8
1
1
With all the complains about heat, its a mobile CPU, they are designed to go up to 100c depending on the model for peak loads. If you dont like it, get a desktop setup or the Framework Desktop, that type of devices can or have better cooling options.

But i was really unhappy with the NVME situation, heat is a much bigger issue here. My two Western 8TB SN850x getting really hot on idle, up to 70c. I now put them in two USB 4 NVME enclosures and temp stay around 35-45c on idle and around 50 on load, yes its just usb 3.2, but 1gb/sec is enough for a home system. Running them in mirrored ZFS for Proxmox LXC Containers.
 

wadup

Active Member
Feb 13, 2024
157
113
43
With all the complains about heat, its a mobile CPU, they are designed to go up to 100c depending on the model for peak loads. If you dont like it, get a desktop setup or the Framework Desktop, that type of devices can or have better cooling options.

But i was really unhappy with the NVME situation, heat is a much bigger issue here. My two Western 8TB SN850x getting really hot on idle, up to 70c. I now put them in two USB 4 NVME enclosures and temp stay around 35-45c on idle and around 50 on load, yes its just usb 3.2, but 1gb/sec is enough for a home system. Running them in mirrored ZFS for Proxmox LXC Containers.
I agree the nvme cooling is really bad which is why I made a 3d printed mount. my drives run under 40C now so hopefully the next one they create a real cooling solutions for nvmes.
 
  • Love
Reactions: pimposh

aaatech

New Member
Aug 11, 2025
2
0
1
Does anyone can change the bios option "PowerLimit Setting"? My bios is 1.02 (F1WSA_FRG_1.02_250616a) and the "PowerLimit Setting" is disabled, I can't change the option, the default value is "Silient Mode", and after I "load optimized defaults", the option change to "Balance Mode", but the menu is gray, I can't click or change the option, I want to change it to "Performance Mode", Does anyone konw how to enable the menu, thanks for help.
 

Attachments

JaxJiang

Member
Jan 10, 2023
88
80
18
Power Limit will be cannot change setting when PCIE GPU inserted.
If you want change your powerlimit setting, you still can change it use manaul fan mode and go to CBS SMU setting.
 

PnoT

Active Member
Mar 1, 2015
661
162
43
Texas
Still no resolution on using the onboard 10gb NICs for a SET team. Can't seem to get help on the subject anywhere so far.