Minisforum MS-01 PCIe Card and RAM Compatibility Thread
- Thread starter Patrick
- Start date
Notice: Page may contain affiliate links for which we may earn a small commission through services like Amazon Affiliates or Skimlinks.
Has anyone tried to bond the interfaces on Debian/Ubuntu and successfully have stability? I have to plug in a 2.5 gbe port and bring down and up the bond for it to resume…
the bond seems to drop network packets daily and make the NICs unresponsive, I can’t even ssh into the host when it starts dropping packets.
i tried 4 different dac cables and 4 different SFP+ ports on the switch.
the bond seems to drop network packets daily and make the NICs unresponsive, I can’t even ssh into the host when it starts dropping packets.
i tried 4 different dac cables and 4 different SFP+ ports on the switch.
Last edited:
Same issue here with a bond of the 2x10 SFP+ interfaces ; not reliable at the moment. Packets randomly dropped.
For you guys having dropped packets, are you getting any errors in the logs ?
What about the switch ?
What about the switch ?
Indeed, I received this e-mail message from Minisforum indicating that they have been on a break:
However, the lack of a Minisforum MS-01 support forum is still an issue. Perhaps they'll get around to configuring their BBS at some point.
Larry
Yes it did, I already received my unit today. It took ~4 days since "Order data transmitted electronically" status to arrive.
I also tried updating the nvm firmware of the NIC to see if it would change something, but nope.
Code:
clariviere@k3s-wg-1:~/700Series/Linux_x64$ sudo ./nvmupdate64e
Intel(R) Ethernet NVM Update Tool
NVMUpdate version 1.40.5.5
Copyright(C) 2013 - 2023 Intel Corporation.
WARNING: To avoid damage to your device, do not stop the update or reboot or power off the system during this update.
Inventory in progress. Please wait [..|*******]
Num Description Ver.(hex) DevId S:B Status
=== ================================== ============ ===== ====== ==============
01) Intel(R) Ethernet Converged 9.64(9.40) 1572 00:002 Up to date
Network Adapter X710
02) Intel(R) Ethernet Controller N/A(N/A) 125C 00:087 Update not
I226-V available
03) Intel(R) Ethernet Controller N/A(N/A) 125B 00:088 Update not
I226-LM available
Tool execution completed with the following status: All operations completed successfully.
Press any key to exit.`dmesg` is not showing anything relevant related to the X710/XL710 or i40e driver.
i'm using a us-xg-16 as my switch with sfp+
Bash:
[ 2807.325508] rbd: rbd3: capacity 1073741824 features 0x101
[ 2807.368084] eth0: renamed from tmp12906
[ 2807.414111] EXT4-fs (rbd3): mounted filesystem f23f945c-e190-4a6e-8983-3c1e9e47ac32 r/w with ordered data mode. Quota mode: none.
[ 2946.846776] libceph: mon1 (2)192.168.1.11:3300 socket closed (con state OPEN)
[ 2946.846795] libceph: mon1 (2)192.168.1.11:3300 session lost, hunting for new mon
[ 2946.847542] libceph: mon2 (2)192.168.1.12:3300 session established
[ 2948.831362] libceph: mon0 (2)192.168.1.12:3300 socket closed (con state OPEN)
[ 2948.831419] libceph: mon2 (2)192.168.1.12:3300 session lost, hunting for new mon
[ 2948.832184] libceph: mon1 (2)192.168.1.5:3300 session established
[ 2961.196366] EXT4-fs (rbd4): unmounting filesystem f21cbd30-245b-4f56-9196-4fc343fbe813.
[ 2961.197988] eth0: renamed from tmp02c57
[ 2961.997759] eth0: renamed from tmp9021b
[ 2962.827197] eth0: renamed from tmpbc3f7
[ 2985.497057] libceph (02700853-57d1-4115-8e57-5458f139a90e e10444): osd3 up
[ 3022.145724] INFO: task jbd2/rbd0-8:8352 blocked for more than 120 seconds.
[ 3022.145740] Tainted: G O 6.5.0-21-generic #21~22.04.1-Ubuntu
[ 3022.145747] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
[ 3022.145752] task:jbd2/rbd0-8 state:D stack:0 pid:8352 ppid:2 flags:0x00004000
[ 3022.145760] Call Trace:
[ 3022.145763] <TASK>
[ 3022.145770] __schedule+0x2cc/0x750
[ 3022.145782] schedule+0x63/0x110
[ 3022.145787] io_schedule+0x46/0x80
[ 3022.145792] bit_wait_io+0x11/0x90
[ 3022.145797] __wait_on_bit+0x4a/0x120
[ 3022.145802] ? __pfx_bit_wait_io+0x10/0x10
[ 3022.145808] out_of_line_wait_on_bit+0x8c/0xb0
[ 3022.145813] ? __pfx_wake_bit_function+0x10/0x10
[ 3022.145821] __wait_on_buffer+0x30/0x50
[ 3022.145829] jbd2_journal_commit_transaction+0x1119/0x19c0
[ 3022.145838] ? lock_timer_base+0x3b/0xe0
[ 3022.145846] kjournald2+0xab/0x280
[ 3022.145852] ? __pfx_autoremove_wake_function+0x10/0x10
[ 3022.145858] ? __pfx_kjournald2+0x10/0x10
[ 3022.145862] kthread+0xef/0x120
[ 3022.145870] ? __pfx_kthread+0x10/0x10
[ 3022.145875] ret_from_fork+0x44/0x70
[ 3022.145883] ? __pfx_kthread+0x10/0x10
[ 3022.145889] ret_from_fork_asm+0x1b/0x30
[ 3022.145899] </TASK>
```
Last edited:
I'm in the same situation. After filing a complaint on February 19th, they sent me a DHL tracking number, which as of today, has not been updated.
some new logs appeared after enabling the option on the switches.
Bash:
[ 7824.652959] Dead loop on virtual device bond0, fix it urgently!
[ 7825.690125] Dead loop on virtual device bond0, fix it urgently!
[ 7827.692661] Dead loop on virtual device bond0, fix it urgently!
[ 7829.969825] Dead loop on virtual device bond0, fix it urgently!
[ 7834.156210] Dead loop on virtual device bond0, fix it urgently!
[ 7838.289398] Dead loop on virtual device bond0, fix it urgently!
[ 7846.443494] Dead loop on virtual device bond0, fix it urgently!
[ 7854.608614] Dead loop on virtual device bond0, fix it urgently!
[ 7860.707074] Dead loop on virtual device bond0, fix it urgently!
[ 7861.680861] Dead loop on virtual device bond0, fix it urgently!
[ 7870.761805] Dead loop on virtual device bond0, fix it urgently!
[ 7886.927047] Dead loop on virtual device bond0, fix it urgently!it didn't seem to kill ceph or such, maybe it just retried and worked
Last edited:
Myne is changed, DHL has got the box and its in Germany export center. I suppose it will leave today or in the weekend and i will have it Monday or Thusday
Now that you mention it, I've checked, and the status has changed, just like yours
he made one for me and it works great. highly recommend.For those of you with Nvidia a2000 someone created a single slot heatsink/fan.
Nvidia RTX A2000 single slot cooler | n3rdware
You can now use your RTX A2000 in extremely small cases like the Lenovo ThinkStation Tiny, Minisforum MS-01 or other cases that only allow low-profile, single slot PCIe cards! Replace the original cooler with this full copper, single slot version and you're good to go.n3rdware.com
a4000
Nvidia RTX 4000 SFF Ada single slot cooler | n3rdware
With this custom full copper cooler, you can use your RTX 4000 SFF Ada in extremely small cases that only allow low-profile, single slot PCIe cards.
He should be releasing soon.
Nice! I emailed him about making a custom one and he said he already made one for someone else and if it worked good he would be selling it. So you were the guy
Do you mind taking some pictures of the finished product? How are the thermals etc
Just FYI this 240w power adapter works perfect with MS-01:
As I noted in an earlier post, these Minis Forum designs (especially the MS-01) are based on certain limitations on power, flexibility and size.
You put in a PCIe card that has to draw large current from the VRM's on the planar, they will generate more internal heat.
If you push the P cores hard, that Alder Lake revision will generate heat and pull more power.
If you stick your MS-01 in some corner or some enclosed rack with limited airflow, you *will* shorten its lifespan.
If you are using both SFP ports, both 2.5 ports and you are using the USB4 ports, regularly or with large amounts (like hours) of high data transfer, it will run very hot and you should have it in a area of reasonable air flow.
The more installed options you employ, the more air flow that MS-01 is going to need.