With a PC connected directly to the modem, can you access 192.168.40.1?
It does seem possible for the ISP to modify the admin IP address via the modem config file, but I have not seen it done. I also cannot come up with a way to find it (not enough coffee, yet), short of a packet capture with your PC directly attached to the modem and hoping the modem advertises its IP address once connected to the ISP.
It is possible that the UK ISP is blocking the port that you have forwarded, for some reason.
As for ping, have you allowed ping on your router, some do not answer ping on their WAN port, by default.
This ^^.
I was having similar problems with a TP-Link C9. Tested with iperf
PC->router->PC, wired
iperf indicated that the C9 topped out at 432 Mbps. Other than a static IP address on the WAN port, the C9 was in the default config.
I’m not sure where the .40. address is coming from. My Arris modem’s admin page lives at 192.168.100.1. I don’t need to do anything to my router to access it.
The router knows that it doesn’t own a 192.168.100.x subnet and forwards that traffic ‘toward’ its default router, where the modem will reply. Not all modems use the .100. subnet. Arris and Motorola do.
In a typical (Arris/Motorola) config, if you watch the ethernet traffic while the modem is coming up and the router is DHCP’ing for its WAN address, you will see the WAN get 192.168.100.xx address until the modem negotiates with the ISP. The modem will then drop and restore link to the router forcing the router to do another DHCP request. The response to this second DHCP request receives the public IP address for the router’s WAN port.
If you run wireshark on a PC connected to the modem while powered off, then power on the modem, you should see a gratuitous ARP advertising the modem’s IP and MAC addresses. This will probably be the management IP address of the modem.
It may be easier to find a sorcerer that can turn the leaden WRT32X into enough gold to buy a pfSense firewall, an ethernet switch and a WiFi6e Access Point.
While the WRT32X does look like a capable xxWRT contender, the networking gods will raise the drawbridge on low latency and high throughput while your packets venture into the dark magical VPN world.
Presuming that you are not subscribed to more than about 500 Mbps, down, a DOCSIS 3.1 modem will not improve much.
If your current modem uses the Intel Puma 5, 6 and maybe 7 chipsets, then a new non-Intel modem may make a difference. If you are not having problems, then a DOCSIS 3.1 modem will probably not improve anything.
If your provider will soon support DOCSIS 4.0, I would wait.
I haven’t done extensive testing, but the Eeros are OK APs when placed in bridge mode. However, that neuters most of the intended functionality and severely limits their configurations. They don’t support PoE and don’t support VLANs.
If you already have them, give them a try. If you are looking to buy them, I’d look at some other WiFi6 AP.
TL;DR, you should be OK.
This article may have some more info. The way that I’m reading the article, the Eeros create a virtual switch fabric and to correctly do that the Eeros need to know what’s connected to each of their ethernet ports and each their WiFis.
What I cannot figure out, is if everything is wire-connected to the first downstream port, will the virtual fabric be correctly set up. I cannot see why not. The Eeros should be able to figure out where everything is (relative to the Eero’s) whether wired or wireless and, thus, forward traffic in the correct direction.
It only seems like there may be a problem if an Eero is not at the entry-point/gateway/router for the rest of the network or the Eeros are all configured in bridge-mode.
To keep it simple (don’t want to cut/terminate my own cable) I’m getting a female/female wall plate.
Does this mean that you want to use a patch cable behind the wall?
If so, make sure it’s CMR (riser/between floors) or CM (not between floors) rated.
Since you have an antenna, I’m assuming that you are getting internet from some form of radio service (WISP or cellular, maybe). What if your family parks somewhere else? Could the car be blocking the signal to the antenna? Does the bandwidth return to normal, if your family turns off their phones?
If you have coax in the two areas, MoCA can provide a wired connection between the router and switch.
If you are strictly using IP addresses in the same subnet, there should not be any name resolution involved. The initiating host will ARP for the MAC address of the host with the IP address that you specified.
If you use ping without the -n option, it will try to resolve the host name associated with the entered IP address. If the DNS query times out or gets an error, ping silently displays the IP address in the host name field instead of the host name. I believe that this is a one and done and should not effect the actual pinging.
You must register the MAC address of the modem with your cable internet provider. If they only allow their modems, you will not be able to use your own modem.
Nothing is wrong with your traceroute to www.google.com.
You have 0% packet loss to the final destination. The mid-path packet loss is due to those routers not responding to ping. Ping is not always a priority.
It would be a better test to traceroute to the actual servers that are problematic.
It depends on the capabilities of your switches. This Cisco article has some foundational info.
Depending on the number of switches, VLANs and trunks, this could get exponentially complicated.
Getting it to work would definitely earn you bragging rights.
Monoprice if you are in the US.
250’ are pretty long ‘patch’ cables. Are the cables in walls or a plenum? Use CMR or CMP rated cables, respectively.
Try moving the cable from port 5 to port 4. If this works, it’s magic.
My reasoning or a really big SWAG, from the pictures, it looks like port 5 might be on a different/singular PHY. The first four ports might be on quad PHY.
It really sounds like a defective switch or a defective power supply, however.
A managed switch IS an unmanaged switch until you configure it.
CCA is more brittle than pure copper and some say will deteriorate over a shorter time.
I use it for some hobbyist stuff (not networking), but wouldn’t pull it through a wall or in a place that I wouldn’t want to go to again.
I don’t think that the cable you referenced is pure copper, since the vendor lists very little about the actual specs. If it were pure copper, it would be listed.
If 8ware is your go to vendor, this is the only bulk cable that I saw, that was listed as pure copper. It’s more than you need, sadly.
As to the purpose of the drain wire: It provides a more reliable ground connection at the plug over only a crimped foil connection.
Modern power supplies are usually of the switching variety and are not line frequency sensitive.
If it’s an old-style analog power supply, a higher line frequency than the rating is OK. It may not be as efficient, but it will work. Using a lower line frequency than the rating will cause additional heating.