CWWK/Topton/... Nxxx quad NIC router
- Thread starter Becks0815
- Start date
Notice: Page may contain affiliate links for which we may earn a small commission through services like Amazon Affiliates or Skimlinks.
Hi everyone, I received a N305 unit a couple of weeks ago that's almost identical to the one that was reviewed on the STH article posted on 08/07/2023. I got it through Amazon but still shipped from China. It has worked well so far. I'm running pfSense on top of Proxmox. The only problem I've encountered appeared after upgrading my fiber connection from 1 gbe to 2 gbe. The tech came yesterday and switched out the fiber converter for a 10 gbe media converter. I have the LAN plugged into a TRENDnet 10G Edgesmart multi-gigabit switch. The issue I've found is that pfSense negotiates the LAN link down to 1 gbe any time I put a load on it. The WAN interface still shows 2.5 gbe link to the fiber converter though. If I unplug/plug the LAN or reboot pfSense the link comes back up as 2.5 again but it doesn't last long. I've tied forcing the link speed by setting the ports on both the switch and pfSense to 2500 but it refuses to connect at all. Has anyone else experienced this type of issue?
I tried changing the port but got the same results. I configured the Eth1 and Eth2 ports as a LAN bridge but got the same behavior. The initial link shows as 2500Base-T then drops to 1000Base-T as soon as any traffic goes across it. I removed the bridge and configured Eth1 and Eth2 as separate subnets. My desktop has a 2.5 gbe and 10 gbe NIC so I unplugged the 10 gbe and plugged the 2.5 gbe directly into the Eth1. I got the full bandwidth on a speed test without any drops. It appears that something is going on with the negotiation of the link speed between the router and both the switches I've tried. The router is the only 2.5 Gbe device that I have plugged into the TRENDnet switch, everything else is 10 Gbe, but it's supposed to handle 1/2.5/5/10gbe. I don't have any issues with the 10 Gbe devices I have plugged in.
I have no problems with 2.5G here, the interfaces are passed through to OPNsense connected to a Ubiquiti USW Enterprise 8 PoE, and even longer iperf3 tests keep the interfaces at 2.5G. Not sure if you can acutally force 2.5G as you did, at least 1G is always auto. If it decides to degrade because of hardware issues (cable quality or length usually), it just does so.
You could run an Ubuntu live OS on USB and see if the interfaces behave the same there. With Proxmox (Debian with Ubuntu Kernel), my management interface also stays at 2.5G, even with longer iperf3 tests.
I think have it working at the moment. I narrowed the issue it down to my TRENDnet switch. I flashed the firmware and restarted it. It has held a 2.5G link since then.
Oof having issues installing opnsense on proxmox following this article How to Install OPNsense in Proxmox in 2023 - WunderTech When i go to start the VM it fails to start with a 192.168.1.200 took too long to respond and the gui is not responsive until a reboot. Bah!
I followed the STH on adding the iommu items and the response was enabled. I did not try zenarmor. i'll do that tomorrow.
Hi @lucker
did you ever get this sorted?
Same here with the fan speeds, they don't do anything, or the opposite. With some strange settings that make no sense at all, I can get the fan to spin at 1200 instead of 900 but that's all. Does not seem to follow temps dynamically.
Hey all, a question for those using proxmox on this kind of devices, how are you monitoring temps?
And how are you checking for the fan spinning speed?
Thanks a lot
And how are you checking for the fan spinning speed?
Thanks a lot
I'm using Zabbix, it's agent has build in support for sensors. CPU and NVME work. Have not found the fan speed sensor yet.
Yes, Zabbix will be for sure something i will deploy further. I was trying to get some info on shell.
Honestly mine isn't an Nxxx device, i've choose a r7-5825u box.
lm-sensors it's actually displayng just this info for the cpu
k10temp-pci-00c3
Adapter: PCI adapter
Tctl: +52.5°C
I was trying to understand if there's a chance to get much detailed info.
Nothing about fan speed here as well
Using stresstestapp with this options:
stressapptest -s 190 -M 256 -m 8 -C 8 -W
I'm getting this sensors' output:
amdgpu-pci-0600
Adapter: PCI adapter
vddgfx: 843.00 mV
vddnb: 743.00 mV
edge: +58.0°C
PPT: 15.00 W
nvme-pci-0500
Adapter: PCI adapter
Composite: +56.9°C (low = -273.1°C, high = +80.8°C)
(crit = +81.8°C)
Sensor 1: +56.9°C (low = -273.1°C, high = +65261.8°C)
Sensor 2: +69.8°C (low = -273.1°C, high = +65261.8°C)
k10temp-pci-00c3
Adapter: PCI adapter
Tctl: +62.5°C
Stressing with s-tui takes the cpu temp in the 64-75° range.
Mostly on 64° once stabilized. But i had to remove the ssd sata, i used as secondary drive to configure the zfs raid1, because only the additional fan or the sata drive can be installed.
Excuse me for polluting this thread and thanks all for the info
Last edited:
I ordered this N100 model from amazon.de and wrote HUNSN to get a bios update. They responded really fast with a link to google drive with a zip file that has the bios and a guide in english. I've reuploaded it on mediafire here, in case the link they sent me was personal and not intended to share.
Could you compare it with the ccwk?
Hi All. was thinking if we perhaps should create a worksheet with which RAM/NVME worked or failed. just placed an order without RAM and NVME as i was hopping to reuse Hardware that i had. but after reading here it looks like a bad idea.. so please if you have the SKU number for the HW used in your box, please please share
RAM
SSDs are WD Black SN750 and Samsung 970 Evo plus
Code:
Manufacturer: Crucial Technology
Serial Number: E73E0D37
Asset Tag: 9876543210
Part Number: CT16G48C40S5.M8A1