mtz_federico

Unfortunately iOS caching can only be done in MacOS. You can find Apple's official guide here: https://support.apple.com/guide/mac-help/set-up-content-caching-on-mac-mchl3b6c3720/mac

Be glad, using this old, unsupported OS online is a ticking security time bomb.

One band has less traffic on it so your ISP steers traffic there or your router picks that signal instead of the other. Nothing out of ordinary. There is nothing wrong.

Router 2 should act as an access point (bridge mode) and not connect as WAN, otherwise devices connected to router 2 will have a separate NAT layer below router 1 and thus devices cannot directly communicate between the two. Huawei ONT Bridge Mode might help

Its absolutely not that simple.

Ping usually uses ICMP, games usually use UDP.  Routers on the Internet can prioritise ICMP to make it appear they aren't congested when being pinged, or they can drop it to lowest priority so that unimportant traffic (such as ping) gets dropped or delayed when under load, rather than useful traffic.

Ping is also handled by the network stack on the server, but gaming traffic is handled by the game server software.  The game server could be completely maxed out, lagging and/or dropping traffic (due to bottlenecks in the software rather than bandwidth), while the server its actually running on has spare capacity to respond to ping like nothing is wrong.  The Ping response might not even come from the server itself, it could come from a router BEFORE the server, a load balancing server, or something doing DDoS protection, it depends on how the network is configured.

Thus even if you know the IP of the game server, pinging it is not necessarily any indication of how it will respond to game traffic.  Only the in-game ping which will most likely be talking to the server software itself can give some indication of the servers health.  But even there its a really small packet of data that is usually only sent every second at most, that in no way reflects the same complexity of actual game traffic, so its a rough guide at best.

Someone who's better at subnet masks can correct me, but wouldn't 255.255.252.0 include both 192.168.1.x and 192.168.2.x?
 
You should probably be using 255.255.255.0 unless you really know what's going on here.

Well if they have Gigabit broadband and are accessing the NAS at the same time they WILL throttle each other.  But like you said, the right fix for that would be to put a 2.5Gbit switch after the router and plug both into that.

Its one reason I upgraded to multi-gig as FTTP is coming soon and I didn't want accessing the NAS to slow downloads on the NAS itself.
 
The second reason being I move files to/from the NAS all the time and didn't want that to impact Plex streaming from the NAS at the same time.

I'd say its definitely worth having at LEAST 2.5Gbit to a NAS, if its half decent even HDDs should come close to that speed for large files.

Yeah that is fine.
 
When it is in router mode it sounds like a wifi issue, are you using 5Ghz? if so, make sure you are not using a DFS channel and that your wifi country is correct (to make sure the allowed bands are used and there is no interference to critical infrastructure).
 
I took a look at an Archer C6 emulator and it looks like when I turn on AP Mode, the DHCP server is still on (this could be why the whole internet goes down). make sure that it is off and on the Network > LAN tab give your router an ip in the same range as your ISP router (use one of the last ones so that it doesn't interfere with another device. Something like 192.168.x.250) and set the Default Gateway to your ISP router's ip. The subnet mask should work at default (255.255.255.0)
 
When you are in AP mode and the internet goes off try switching the cable from WAN to LAN (or the other way around).

I would leave the ISP router as default (make sure it is not in bridge mode and the DHCP server is on) is and setup the second router (Router A) in AP mode. You mentioned that it is a TP-Link router in AP Mode, AP mode should disable the DHCP server (that's probably why you can't see it). I have only used old TP-Link routers that don't have AP Mode and have had to do it manually but from what I've seen, it should just work. Just make sure that the connection from the ISP router is connected to the "WAN" port (it is usually yellow)
 
ipv4 can only be automatically assigned to devices via DHCP while ipv6 is usually assigned with Router Advertisements while can also use DHCPv6 but is rarely used (at least in home setups) because Android refuses to add support for it. Take a look at the ipv6 settings on both the ISP router and on the TP-Link, the TP-Link should not be doing anything besides letting the RA or DHCPv6 work.
 
How is this happening? that is, are your devices disconnected from the wifi, does the internet just stop working, do you get an error message?

I would leave the ISP router as default (make sure it is not in bridge mode and the DHCP server is on) is and setup the second router (Router A) in AP mode. You mentioned that it is a TP-Link router in AP Mode, AP mode should disable the DHCP server (that's probably why you can't see it). I have only used old TP-Link routers that don't have AP Mode and have had to do it manually but from what I've seen, it should just work. Just make sure that the connection from the ISP router is connected to the "WAN" port (it is usually yellow)
 
ipv4 can only be automatically assigned to devices via DHCP while ipv6 is usually assigned with Router Advertisements while can also use DHCPv6 but is rarely used (at least in home setups) because Android refuses to add support for it. Take a look at the ipv6 settings on both the ISP router and on the TP-Link, the TP-Link should not be doing anything besides letting the RA or DHCPv6 work.
 
How is this happening? that is, are your devices disconnected from the wifi, does the internet just stop working, do you get an error message?

Yes, you do need IPv6. CGNAT is a huge deal and causes tons of problems that IPv6 gets around. It has NOTHING to do with number of devices for home users.

I'm building a network setup for my cabin, using a Att mobile hotspot into a Tp-link TL-SG105 to run fire sticks. Running an antenna to the roof to boost signal. Mounted it all onto a live edge board. 

Is there a question or just sharing?

These are limits of the ethernet spec, so all deviecs have the same 100M limit. You won't have any issues here with these run lengths.
 
You ports will connect at 100m/1g only, so if you have a 800m link from the isp, its running at a 1g link speed.
 
 

That's precisely what the reverse proxy would do.
You get a domain, you make different hosts at your domain provider that are e.g. mc1.yourdomain, mc2.yourdomain etc.
The reverse proxy is configured so that if whoever tries to connect gave mc1.yourdomain:25565 then that is redirected to port X, if they gave mc2.yourdomain:25565 it's redirected to port Y. You have to configure your different servers run on port X and Y internally.

Boot a bootable Linux like Ubuntu and check if the ethernet works. If it does, then the OS is the issue.

They’re probably not the same fiber cable, but they are probably in the same bundle. Fiber main lines are often 96 fibers (48 pairs), these then go off into buildings or a small area as usually 6 pairs or 12 pairs. Each fiber pair can be leased to a different customer. So Spectrum may have run the fiber bundle to your building that everyone is leasing, one pair is being used for the cable internet service while another pair is being leased by T-Mobile. Usually when there is fiber damage, the whole bundle needs to be re-spliced. Even most of the bundle survived, they may need to cut it and put in extensions to complete the full repair. A 96-fiber repair takes 4-8 hours, *after* the highly trained crew makes it to the site and clears sufficient access to the bundle. Ask your friend whether drivers or backhoes cause more issues for him - most of the outages that affect the company I work for are caused by backhoes or other construction equipment. Especially when they skip paying (or waiting) for a buried cable and pipe survey.
 
For true “cell towers” they often have fiber (as mentioned by others, there is a difference between an enterprise dedicated fiber connection, and what is deployed by FiOS and other fiber-to-the-home) as a main connection, and microwave point-to-point connections to other nearby towers as a backup. If the tower is in a place that’s expensive to reach, it might only have microwave connections. On a cell tower, you can spot the microwave connections by looking for a kick drum.
 
In your building (you didn’t really confirm or deny whether you live in an apartment building, I’m going to assume you do), you probably don’t have a “cell tower”, and your building definitely doesn’t have any backup connections, at least for T-Mobile. Many buildings have “micro nodes” or DAS “distributed antenna system” which means that an indoor version of the antennas on a cell tower are placed on every floor or every other floor. The antennas connect using coaxial cable (same material as the “coax cable” for a cable internet or TV service, but different signals put onto it) to a headend device, which can take fiber or ethernet connections. The headend device is usually in the same equipment room as the main cable tv/internet splitter/amplifier/etc.
 
In my company’s HQ, we have a DAS for AT&T which is using fiber from Cox (a “cable” company), and a DAS for Verizon Wireless which is using fiber from Verizon (the “phone” company, which is legally separate from Verizon Wireless). Neither DAS has a backup connection. The company also independently purchases dedicated fiber services from Verizon and Lumen, as well as cable modem internet and phone services from Cox. Although Cox fiber comes into our building, the “cable headend” that is converting from fiber to coax is located outside in a pedestal, because it services the whole block. My HQ is more complicated than any apartment complex, but the way the DAS’ are connected is relevant.

If you live in a moderately dense apartment building or similar, there is probably a local cell tower on or in your building, that is connected via fiber probably to spectrum. Or even if its a separate company connecting the cell tower, the fibers go through the same path that the drunk person hits. The cell tower should stop broadcasting when it is offline, but it might not.

When I said niche feature, I was referring to VLAN tagging.
But anyway, I strongly doubt that it was a CPU issue that prevented you from getting more than 500Mbps throughput. 
 
Besides, the routers shown in the video are quad cores at 1.8GHz, and a far superior architecture to the one you have in your router. So we're talking like 20 times the performance of your router. On top of that, some workloads are not even running on the CPU. I can't comment on the specific processor your router uses, but the processor in the routers mentioned in the video both have the BCM49408 which has hardware accelerated circuitry for NAT, on top of a dedicated processor for the wireless portion. So things like NAT does not run on the CPU at all, with the exception of maybe under some circumstances with some specific features enabled (like certain logging or inspection features). Neither does any of the wireless processing.
 
 
Home routers are very purpose built, and very good at what they were designed to do. Most of the information in this video is bollocks because it does not apply to the 99% of people. The CPU being "bad" does not matter for home routers because the CPU does very little to begin with, and the few things they do do not require a lot of processing. 
It's when you start enabling some features that might break certain hardware acceleration that you may run into problems. But most people do not do that so it's not an issue.

You would need to run a PBX server. Crosstalk sollutions has some good videos on how to set this up. I have never set one up. If you need help, he has a discord server.

pfsense is a router os. It does not use docsis. Your modem uses docsis to convert your internet connection from coaxial cable to ethernet. pfsense just takes that ethernet connection. the issue seems to be with DHCP, the weird thing is that your computers are getting a connection but not the routers.
 
For testing I would try to set a manual ip on pfsense. Connect the modem to your computer, write down the ip, gateway and subnet mask and set them on pfsense. This rules out the actual connection and confirms that the issue is with DHCP.
 
I am asuming that both the AMD desktop and the HP laptop are running windows. I would tell spectrum that the direct modem (wan connection) is working on windows but not on any router.

I hope these numbers are just "for example" because these speeds sounds like DSL-era type of speeds.
 
Wire colour is not helpful at all.
 
This sounds like something only the ISP can help with. Contact them so they can send a technician out to troubleshoot the issue. Quote the speeds you're paying for versus what you're getting over ethernet. Test speeds again before the technician leaves.

For me cloudflare (1.1.1.1, 1.0.0.1) is the fastest but I like to use quad9 (9.9.9.9, 149.112.112.112) because It offers malware protection and the speeds stop making a difference after a certain point.

For the router WAN and switch lan1 connection, you have a choice. Either you set lan1 to VLAN40 untagged (aka access), and then on the router it doesn't need to be aware of any VLAN at all (the packets between it and the switch won't have any VLAN tag, hence "untagged"), or you set *both* the lan1 and wan ports to use tagged aka trunk with VLAN 40 (the packets between the switch and the router will have the VLAN tags on them). Since you only need to keep VLAN separation on the switch itself, I would prefer to set the lan1 port to untagged/access.

I would love to see USB-C to Display Port Cables tested. I feel like those cables are much more likely to be unreliable.