I'm seeing the same with my own 4-port CWWK N305 with near identical components.
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.
Looks good. How did you add those monitoring widgets?
EDIT: I am thinking it comes from something else than proxmox, HA or something.
I have questions:
I put the H piece and first I put my nvme in the H piece (AKA in the center). Installed everything and forgot about it. Then I opened the box to show it to my friend and realized that that was the x1 port and I swapped it to the other nvme port.
Machine boots but for some reason my internet connection didn't work at all. Couldn't ping anything (ip or dns).
Fastforward, I re-installed proxmox (I wanted to do it anyway) and everything good again but the `lspci` still says x1
I tried a hdparm and I got:
Timing cached reads: 23070 MB in 2.00 seconds = 11550.59 MB/sec
Timing buffered disk reads: 2492 MB in 3.00 seconds = 830.55 MB/sec
----
Not sure about the results.
What should Iook next?I saw images here where it says x4.
My bios is from september. Maybe I should update it (I haven't touch anything on it yet).
---
I tried again the other nvme (the one in the H) and says the same x1. Out of curiosity, booting it in the different port made my proxmox lose internet connection again (but enought to check lspci). Put it back and works again. No idea why changing the port messes with internet.
I put the H piece and first I put my nvme in the H piece (AKA in the center). Installed everything and forgot about it. Then I opened the box to show it to my friend and realized that that was the x1 port and I swapped it to the other nvme port.
Machine boots but for some reason my internet connection didn't work at all. Couldn't ping anything (ip or dns).
Fastforward, I re-installed proxmox (I wanted to do it anyway) and everything good again but the `lspci` still says x1
I tried a hdparm and I got:
Timing cached reads: 23070 MB in 2.00 seconds = 11550.59 MB/sec
Timing buffered disk reads: 2492 MB in 3.00 seconds = 830.55 MB/sec
----
Not sure about the results.
What should Iook next?I saw images here where it says x4.
My bios is from september. Maybe I should update it (I haven't touch anything on it yet).
---
I tried again the other nvme (the one in the H) and says the same x1. Out of curiosity, booting it in the different port made my proxmox lose internet connection again (but enought to check lspci). Put it back and works again. No idea why changing the port messes with internet.
So let me summarize:I have questions:
I put the H piece and first I put my nvme in the H piece (AKA in the center). Installed everything and forgot about it. Then I opened the box to show it to my friend and realized that that was the x1 port and I swapped it to the other nvme port.
Machine boots but for some reason my internet connection didn't work at all. Couldn't ping anything (ip or dns).
Fastforward, I re-installed proxmox (I wanted to do it anyway) and everything good again but the `lspci` still says x1
I tried a hdparm and I got:
Timing cached reads: 23070 MB in 2.00 seconds = 11550.59 MB/sec
Timing buffered disk reads: 2492 MB in 3.00 seconds = 830.55 MB/sec
----
Not sure about the results.
What should Iook next?I saw images here where it says x4.
My bios is from september. Maybe I should update it (I haven't touch anything on it yet).
---
I tried again the other nvme (the one in the H) and says the same x1. Out of curiosity, booting it in the different port made my proxmox lose internet connection again (but enought to check lspci). Put it back and works again. No idea why changing the port messes with internet.
- SSD in native 2280 slot: network problems
- SSD in H-bracket: OK
This would add another test for me, because my SSD is also in the native 2280 slot and I'm also facing network problems.
Nah, not really. It is just that after installing in one slot, if I move the nvme drive it boots but there is no internet connection. Happened in both slots if I swap it.
----
Put opnsense (through proxmox) and I didn't manage to get it working with PPPoE. PfSense works like a charm. Why? No idea.
Question (Or I can ask elsewhere):
NIC1: Proxmox
NIC2: WAN Pfsense
NIC3: LAN Pfsense (actually it is just my computer)
I pass WAN directly as PCI and LAN as vmbr1. Vmbr0 is the proxmox one with static ip address. In pfsense I have a network 10.12.6.1/24 and DHCP enabled so my computer will get one IP and work. Everything works like a charm.
How can I access proxmox within my PFsense network? Can I do that? Without a physical cable in NIC1. Even if I put vbmr0 static ip to be like 10.12.6.2 I can't access it. If I swap my cable, yes, it works, I can access proxmox but not pfesense (or internet for that matter).
Out of curiosity, if I am at NIC1 (proxmox, no internet, no pfsense) and I create a VM where I bridge vmbr1 (LAN) I can access to internet through that VM but still not in proxmox or my computer.
Is it there a way or I need extra hardware (like a switch).
Thanks.
Changing the pcie port might change the numbering of your pcie devices including NICs, no?
About my issue with nics, proxmox and pfsense. I ended putting a static ip and gateway to vmbr1 (LAN) so if my network is 10.12.6.1/24 I put 10.12.6.2/24 as the IP so I have everything I need. I can connect to proxmox, have internet and if I shutdown pfsense I still access to proxmox.
EDIT WITH FIX: Sorry this became a monologue but I will leave here my original problem and the solution in case anyone encounters this in the future.
The shell that runs with the bios defaults to fd0 (which can be your NVME) and tries to execute everything from your NVME. That is why they suggest removing it before hand unless you know how to change to fd1 and execute the steps by hand.
------
Original issue: I think I know where my x1 downgraded issue comes from. I ordered the board for 4 nvme and they flashed the bios. I just wanted to have it "just in case".
I tried to download the bios named:
CW-AL-4L-V2.0(先锋四网N100-N200-I3-N305-V2出厂默认原始版本)2023.09.27
AKA
CW-AL-4L-V2.0 (Pioneer Quad-Network N100-N200-I3-N305-V2 Factory Default Original Version) 2023.09.27
I flashed it with rufus, booted into a shell and says something about `1.nsh` not working out. I see the script in the ISO but...
I saw a video of installing Ventus to load the ISO within the machine but same result.
The shell that runs with the bios defaults to fd0 (which can be your NVME) and tries to execute everything from your NVME. That is why they suggest removing it before hand unless you know how to change to fd1 and execute the steps by hand.
------
Original issue: I think I know where my x1 downgraded issue comes from. I ordered the board for 4 nvme and they flashed the bios. I just wanted to have it "just in case".
I tried to download the bios named:
CW-AL-4L-V2.0(先锋四网N100-N200-I3-N305-V2出厂默认原始版本)2023.09.27
AKA
CW-AL-4L-V2.0 (Pioneer Quad-Network N100-N200-I3-N305-V2 Factory Default Original Version) 2023.09.27
I flashed it with rufus, booted into a shell and says something about `1.nsh` not working out. I see the script in the ISO but...
I saw a video of installing Ventus to load the ISO within the machine but same result.
Last edited:
well, FYI, I made overview video about CWWK 2 port version
We also may want to consider the TDP in the bios of these boxes. The N305 in the 4x2.5ghz mini pci i have was set to use default TDP (15 watts) under short bursts, but then the Package is set to down the TDP to 10watts under load. What this translated to was a max all clock was 1.8ghz when all cores were being pushed.
I played with the TDP of the CPU and Package yesterday a bit and found a few things. For one you can set the Package and the cpu above 15 watts and it will run though i wouldn't do it for long. I was able to do that and found that i could get my minipc to pull down over 40 watts and could run all cores at over 3ghz with the heat still below throttling. I hope i didn't create any damage because when it got to the cpu heat produced when 1 core was being pushed it would push the temps close to throttling.
What I ended up doing was simply setting the down TDP setting to 17.5 watts instead of 10 watts and that allowed me to maintain 2.3ghz on all cores and the heat isn't to bad. That isn't a bad increase from 1800. May be a way to squeze a bit more performance out of it. The low power usage is still the same though.
I played with the TDP of the CPU and Package yesterday a bit and found a few things. For one you can set the Package and the cpu above 15 watts and it will run though i wouldn't do it for long. I was able to do that and found that i could get my minipc to pull down over 40 watts and could run all cores at over 3ghz with the heat still below throttling. I hope i didn't create any damage because when it got to the cpu heat produced when 1 core was being pushed it would push the temps close to throttling.
What I ended up doing was simply setting the down TDP setting to 17.5 watts instead of 10 watts and that allowed me to maintain 2.3ghz on all cores and the heat isn't to bad. That isn't a bad increase from 1800. May be a way to squeze a bit more performance out of it. The low power usage is still the same though.
Curious how 2 port users get on with that set up since software bridging is it that bad?
Are more people relying on wifi for the rest of their home devices or using smart switches or something else perhaps.
I tend to stick with good old hard wired ethernet on my computers and few other devices around home, so 6 ethernet cwwk units I prefer.
Are more people relying on wifi for the rest of their home devices or using smart switches or something else perhaps.
I tend to stick with good old hard wired ethernet on my computers and few other devices around home, so 6 ethernet cwwk units I prefer.
I used 2 port version because its all what you need, you don't bridge interfaces to make switch*
One is ethernet input, one is ethernet output and thats all, proxmox things will be separated using vlans (on output port)
this hardware is not replacing router/switch, it can act just a firewall (pfsense or opnsense) and of course proxmox goodies...
bigger amount of ethernet ports is just convenient if you want to split it for separated vlans and that's all.
So software bridging is bad in scenario where your n100 is acting like a switch, for switching there are switches with their own ASICs
*well, in reality in proxmox you would anyway pass bridged interface to pfsense/opnsense virtual machine, you could pass pcie NIC without bridge for WAN for example and it will be same/fine but for LAN port you would use software bridge anyway because other vms/lxcs must have linked network connection to a LAN interface -software bridge/switch-
ahh ok makes sorta makes sense, I have a ton of networking still to learn especially with vlans, am just using openwrt with 6 port nic router with bridged setup, nothing fancy like proxmox/vm set ups just running a wireguard vpn so all devices on the home network are secure and policy based routing for devices that need clear net.
Hi, Does anyone can repost this BIOS from the manufacturer. Provided Link is Dead.
I might want to put a fan on this. I see the solution of putting a 80mm on the bottom (outside) but that would force the computer to be upside down but more importantly to have a naked fan on top that can collect dust in a matter of days.
Putting a 120mm on top is well, the same.
I might need to re-read this thread to see what solutions you put. But before that I will change the paste today and put a small heatsink on the nvme
Putting a 120mm on top is well, the same.
I might need to re-read this thread to see what solutions you put. But before that I will change the paste today and put a small heatsink on the nvme
another solution is to put small fan inside case next to ssd and ram. I have asked crucial for ram operating temperatures and they replied me up to 85 celsius degrees
But with SSD I would like to stay under 60.Buw how small? There is not much clearance. 10-15mm I would say.another solution is to put small fan inside case next to ssd and ram. I have asked crucial for ram operating temperatures and they replied me up to 85 celsius degrees
No idea how is my ram doing, lm-sensors doesn't output it (I am open to other ways to check).
My NVME is normally around 50. Now it is 60 because I did a stress test and I guess it will take a while to go back down since the case is closed.
I will put a heatsink as mentioned and re-paste.
I also saw a 3d printed stand for a similar model (easy to modify to fit our needs). I can put the machine on top of the stand and there is enough space for a 25mm fan on the bottom. Will look cool and less temperature. We will see. I will keep you all updated.