-1

I have a friend who has a setup with two ISP's. They are running an insurance agency and are obliged to use an ISP from their insurance company. Problem with this ISP is that it is superslow (4Mbit) and it has many website and port restrictions.

I had made a setup for them in the past, where they would use a High Speed ISP for all their basic internet needs and everything that needed to be done on the insurance ISP was routed to the Insurance router.

Basically the setup was:

Router A (Cable Internet HIGH Speed ISP) : IP: 192.6.1.1/24 DHCP: ON Gateway

Router B (Insurance company ISP) IP: 192.6.1.120/24 DHCP: OFF

On router A, i've set up two routing rules for all trafic meant for the Insurance Isp networks which would be routed to the IP of Router B

This setup has worked great for many years.

But last fall they merged with another agency (same Insurance company though) and they changed office (actually they merged office with the other agency). They wanted to set up the same kind of network setup on this new location. There was already an office set up on this location with a router from the Insurance ISP. Insurance ISP Router at this office was set up 192.168.1.1/24.

They also had a couple of computers and printers setup with fixed ip's so instead of changing everything i disabled DHCP on the Insurance ISP router and changed the Router IP of the High Speed ISP. I've also adjusted routing tables to redirect all Insurance Ip traffic to 192.168.1.1.

So current setup is:

Router A (Cable Internet HIGH Speed ISP) : IP: 192.168.1.254/24 DHCP: ON Gateway

Router B (Insurance company ISP) IP: 192.168.1.1/24 DHCP: OFF

This works for a couple of days/weeks and then suddenly their network starts slowing down to a less than 1Mbit internet speed. If i do a full reset of both routers and attached switch i can get everything running again at full speed for a couple of days/weeks but then again after some time internet speed will fall back to unacceptable speed of less than 1Mbit. How come this worked flawlessly for many years at location1, but starts stalling at Location2 after just a couple of days/weeks?

Is there anything i can change at the setup for location B? (someone suggested me to put router A at 192.168.1.1 and put router B at a higher IP, but would this actually solve anything? I need to contact the insurance ISP to do this, since i have no access rights on their router, it was already quite a hassle to make them turn off DHCP, i'd rather not call them to adjust the router ip if it's not necessary)

EDIT: extra info

Router A:

  • model Cisco RV130
  • (ip: 192.168.1.254, subnet 255.255.255.0, gateway)
  • DHCP: ON (DHCP range 192.168.1.50 to 192.168.1.150)
  • WAN port: connected to Consumer High Speed ISP modem
  • LAN port 1: Connected to linkys switch LGS116 (unmanaged)
  • LAN port 2: Connected to Router B
  • No other lan ports are connected

following routes are set up:

  • 172.31.0.0/16 to 192.168.1.1
  • 212.19.64.0/19 to 192.168.1.1

Router B:

  • Model Cisco 891 (managed by insurance ISP)
  • (ip: 192.168.1.1, subnet 255.255.255.0)
  • DHCP: OFF
  • Wan port: connected to Insurance ISP (4Mbit)
  • LAN port 1 connected to Lan port 2 on Router A.
  • No other lan ports are connected

Strange thing is that this setup worked for years at location1; But at location2 it doesn't, it only works for a few days/weeks.

Router B is the same model at both locations, but they are two different routers at both locations. Router A is the same router which we moved from location1 to location2. Swithch is also the same switch which we moved from location1 to location2. Location2 has some 4port desktop switches (Netgear GS105) in the office, which is different to location1.

What i meant with reset is a power cycle, and this brings back the network at full desired speed. For the moment, I was planning to do a nightly power cycle with a programmable power socket, but IMO that's not a real solution, more a bypass. I would like to know what's wrong here and why this worked without a problem at location1 for 3-4 years and now at location2 starts slowing down to unacceptable speed after some time (mostly at Friday noon)?

I have full admin rights on router A, router B is managed by the insurance company ISP.

network diagram

closed as too broad by bertieb, harrymc, LPChip, Tim_Stewart, Burgi Mar 6 at 9:10

Please edit the question to limit it to a specific problem with enough detail to identify an adequate answer. Avoid asking multiple distinct questions at once. See the How to Ask page for help clarifying this question. If this question can be reworded to fit the rules in the help center, please edit the question.

  • If a reboot works (if that's what you mean with the "full reset"), then could just have the routers reboot themselves weekly – Xen2050 Mar 2 at 12:25
  • 2
    You neglected to advise the make & model of the routers, and method of connection, nor have you described which interface things connect to or where NAT is being done, so we cant help you. – davidgo Mar 2 at 17:23
  • Welcome to super user, this question seems to better suited for the server fault community. But I suspect without more information in your question it will be put on hold over there. – Tim_Stewart Mar 2 at 17:25
  • 2
    Its difficult to say for sure, but it does sound like the issue is because both routers use the same subnet (maybe caused by ARP;ICMP redirects) Changing the subnet could fix this. In any event, changing one of the network ranges will facilitate better control of the network. – davidgo Mar 2 at 17:37
  • 1
    If only you can link a sketch the network topology/addressing/configuration so that we can help you – Soulimane Mammar Mar 3 at 15:08
0

I can think of 2 possible issues. The first thing I would check is that there is no routing loop, ie that there is only 1 connection from each switch to another (and im including the LAN ports on the router as a switch here). You may be able to test if this is the issue by leaving the routers on but rebooting all the switches next time the slowdown occurs.

Less likely, but still worth considering is a router bug, possibly related to ICMP redirects. I am a control freak so I would change the IP range between router A and B - which will also require a route to be added on router B to push traffic destined for your LAN via router A. Another possibility (which Im hesitant to suggest, but might work) is to check if ICMP redirects are being generated by router B and disable them if they are.

Something which may be useful to you in thinking about networks/routers is that IP addresses are associated with Interfaces rather then devices. This is relevant as Im pretty sure router A only has network interfaces and both router A and B are on the same LAN (which is not ideal, but not uncommon), even though your diagram shows them as being on seperate LANS/segments.

  • David, thanks for your input. To make things clear, both routers are indeed on the same LAN; I'm not a networking expert and the diagram is something i've created today, but i don't really know the rules for designing such diagrams. Should i move router B lower (on the same level as the first switch) to clarify? Routing loop was also the first thing i've checked. When this slowdown happened for the first time, i've disconnected the main switch (linksys) and did a check on a free lan port of Router A. But even then the slowdown is still there. – Tom AC Mar 3 at 21:09
  • Its probably not a big deal. If you need this diagram as a network reference I'd be inclined to join the 2 lines on LAN side of the Cisco RV130 router, which (although mixing layers) shows they are both on the same layer 2. A bit difficult, but I'd also be inclined to mark IP addresses outside the devices near the interface, but its not a big deal and easy enough to work out once we know the RV130 only has 1 LAN Interface. I don't profess to be professional at doing network diagrams. – davidgo Mar 3 at 21:17
  • Changing the Ip-range between router a and b is not possible, because of the route that needs to be added on Router B; I don't manage this router, the insurance company ISP does and they are not really keen on making modifications on this setup. We got them to disable DHCP by telling there is a Windows SMB server on the network which needs to be DHCP (and there was such a server in the past). I can maybe convince them to change the IP on router B but i'm afraid that's as far as it gets. – Tom AC Mar 3 at 21:29
  • From your last comment it sounds to me like the problem is with your RV130 router (and indeed Amazon reviews for this router for say it is riddled with bugs - which does not surprise me as its not even a "real" Cisco router as it runs Linux). You might be better off getting yourself another mid-range router which runs DD-WRT (I use ASUS RT-AC68U in places like this), flash DD-WRT on it - this will be more stable. As a bonus you can then turn one of the LAN ports into a second WAN port and plug the 891 router into that, which is the "proper" way of doing things. – davidgo Mar 3 at 21:30
  • You can always change the range between router A and your clients. I'd recommend using 10.0.1.x netmask 255.255.255.0 - Hassle to redo things like printers and servers though. A new router A where you can have multiple WAN interfaces is also a really good idea if dealing with a hard-to-deal-with provider like the insurance company. – davidgo Mar 3 at 21:31

Not the answer you're looking for? Browse other questions tagged or ask your own question.