2 Routers connected to 1 Cable modem?

[randomName]

Hi there, again. So I've been consistently seeing better performance from my router when all devices are disconnected from it besides one. Obviously. So it got me thinking if it's possible to connect 2 routers directly to 1 modem. My MB8600 has 4 ethernet ports but I think it's for Link Aggregation. Aside from hardware obstacles is it possible to do this? The goal is to limit connections from one router to one device while the other handles the rest of my devices in the house.

So is it possible?

Thanks for you time

 

[L&LD]

Hi there, again. So I've been consistently seeing better performance from my router when all devices are disconnected from it besides one. Obviously. So it got me thinking if it's possible to connect 2 routers directly to 1 modem. My MB8600 has 4 ethernet ports but I think it's for Link Aggregation. Aside from hardware obstacles is it possible to do this? The goal is to limit connections from one router to one device while the other handles the rest of my devices in the house.

So is it possible?

Thanks for you time

This would mostly depend on how many IP's your ISP gives you. Most give only one. But if you are lucky to be allowed two?

If two, then yes, this should work.

But, how many devices are connected to your network that you notice it slowing down? Or, is this a case of needing higher bandwidth (up/down or both?), in which case, two routers won't help here, as they'll be fighting each other to give the most voracious clients the most bandwidth they can.

 

[randomName]

This would mostly depend on how many IP's your ISP gives you. Most give only one. But if you are lucky to be allowed two?

If two, then yes, this should work.

But, how many devices are connected to your network that you notice it slowing down? Or, is this a case of needing higher bandwidth (up/down or both?), in which case, two routers won't help here, as they'll be fighting each other to give the most voracious clients the most bandwidth they can.

The range is very narrow but noticeable. I browse the internet on my main rig while gaming on the other. Those are the only two wired connections I have then there are other devices like my cell phone that will show connections but other than those devices active while gaming I have 2 smart TV's in the house but they aren't one while I'm at my computer.

I did talk to a tech about it while calling for another reason but I don't think he had indepth knowledge of how to set it up. He just said the only way would be to put a switch between the cable modem and hook both routers to that, which I would assume would defeat the purpose. Never mentioned how many IPs they allow.

 

[L&LD]

The range is very narrow but noticeable. I browse the internet on my main rig while gaming on the other. Those are the only two wired connections I have then there are other devices like my cell phone that will show connections but other than those devices active while gaming I have 2 smart TV's in the house but they aren't one while I'm at my computer.

I did talk to a tech about it while calling for another reason but I don't think he had indepth knowledge of how to set it up. He just said the only way would be to put a switch between the cable modem and hook both routers to that, which I would assume would defeat the purpose. Never mentioned how many IPs they allow.

What is your ISP paid-for-speeds? Do you reach or exceed them on a regular basis during the course of a day?

Even if you needed to use a switch between your modem and two routers, you will still need to be connected to two different IP's on their end. But you will be sharing the connection between the router, even then.

I say 'even if you needed a switch' because your modem has 4 ports on it, correct? If they allowed you to have two or more public IP's, then at least two ports would be 'live' on your modem. That is something you can test now.

 

[randomName]

What is your ISP paid-for-speeds? Do you reach or exceed them on a regular basis during the course of a day?

Even if you needed to use a switch between your modem and two routers, you will still need to be connected to two different IP's on their end. But you will be sharing the connection between the router, even then.

I say 'even if you needed a switch' because your modem has 4 ports on it, correct? If they allowed you to have two or more public IP's, then at least two ports would be 'live' on your modem. That is something you can test now.

30/3, not that I'm aware of.

Yes the MB8600 has 4 ports and I did mention that to the tech but I didn't get the idea that he knew how to answer that.

 

[coxhaus]

If you want you can connect a switch to the modem then plug all your routers into the switch using the WAN port. This works real well if you have 5 static IPs. All the routers will be assigned an IP until you run out of ISP IPs.

 

[randomName]

If you want you can connect a switch to the modem then plug all your routers into the switch using the WAN port. This works real well if you have 5 static IPs. All the router will be assigned an IP until you run out of ISP IPs.

Wouldn't the method of using a switch act the same way as a router in terms of connections?

EDIT: I would rather not use a switch if I didn't have to.

 

[L&LD]

30/3, not that I'm aware of.

Yes the MB8600 has 4 ports and I did mention that to the tech but I didn't get the idea that he knew how to answer that.

With your provided ISP speeds, I can see how you will notice when you are the only client on the network and when others are sharing it. It's the upload that is the real issue for you, I would guess.

I believe you have the FreshJR QOS script installed, have you optimized your settings in the Adaptive QoS page? You should have it set to Adaptive QoS, fq_codel selected, and your Manually entered up and down speeds in the 85% to 95% range of your actual speeds.

Using dslreports.com/speedtest and selecting your connection type to perform the test, you want to fine-tune the up and down speeds to have the highest possible speeds with the lowest possible bufferbloat (yes, it is a balancing act).

With 30/3 Down/Up, I would start at 30/2.5 and go slowly up or down from there (for dialing in the upload speeds). Remember that you need to wait at least 10 minutes after making adjustments to the QoS page for them to apply properly. Also, keep in mind that for gaming, the upload is usually more important than the download (bufferbloat wise).

 

[coxhaus]

Wouldn't the method of using a switch act the same way as a router in terms of connections?

EDIT: I would rather not use a switch if I didn't have to.

No. The switch will be layer 2. The router is layer 3.

 

[randomName]

With your provided ISP speeds, I can see how you will notice when you are the only client on the network and when others are sharing it. It's the upload that is the real issue for you, I would guess.

I believe you have the FreshJR QOS script installed, have you optimized your settings in the Adaptive QoS page? You should have it set to Adaptive QoS, fq_codel selected, and your Manually entered up and down speeds in the 85% to 95% range of your actual speeds.

Using dslreports.com/speedtest and selecting your connection type to perform the test, you want to fine-tune the up and down speeds to have the highest possible speeds with the lowest possible bufferbloat (yes, it is a balancing act).

With 30/3 Down/Up, I would start at 30/2.5 and go slowly up or down from there (for dialing in the upload speeds). Remember that you need to wait at least 10 minutes after making adjustments to the QoS page for them to apply properly. Also, keep in mind that for gaming, the upload is usually more important than the download (bufferbloat wise).

Yes I managed to successfully get it installed

I think so. I have it set to Adaptive QoS, fq_codel, and manually entered 80% of 30/3 which is 24/2.4 I forgot what it was before I manually set it; can I just put it both at zero for testing at dslreports.com/speedtest?

 

[L&LD]

Wouldn't the method of using a switch act the same way as a router in terms of connections?

EDIT: I would rather not use a switch if I didn't have to.

The switch acts in a many to many (connection) configuration. The router acts as a single to many (connection) configuration for the LAN clients on it.

In the case of the switch used between the modem and multiple routers, it is giving multiple public IP addresses to multiple routers. If, the ISP allows for multiple public IP's.

Did you have a chance to test if more than two ports are 'live' on your 4 Port modem? All you should need to do is move your WAN connected cable from the port it is connected to now on the modem, to another one.

You may also need to power down your modem, wait at least 10 minutes and then power it back up for the ISP to 'see' your connection on the other Port used.

If you can test all the ports in this fashion, and at least two of them are 'live', then borrow or otherwise test with an additional router with both ports concurrently. If both routers get a public IP and continue to operate normally, great!

Or, you can try to call/talk to a more knowledgable employee on your ISP's staff and ask them directly what your contract allows you to do, regarding multiple public IP's (whether static or dynamic, shouldn't matter for you).

 

[L&LD]

Yes I managed to successfully get it installed

I think so. I have it set to Adaptive QoS, fq_codel, and manually entered 80% of 30/3 which is 24/2.4 I forgot what it was before I manually set it; can I just put it both at zero for testing at dslreports.com/speedtest?

No, you don't want them set at zero.

Test what you have set them to now, note the dslreports.com results and then change them slightly and see if there is a positive difference or not. Slowly adjust your guesses until the numbers provide you with the greatest download speeds with the least (lowest) bufferbloat.

 

[randomName]

No, you don't want them set at zero.

Test what you have set them to now, note the dslreports.com results and then change them slightly and see if there is a positive difference or not. Slowly adjust your guesses until the numbers provide you with the greatest download speeds with the least (lowest) bufferbloat.

For getting a base reading of my ISP speed I turned off Adaptive QoS. Ran a few test from Speedtest.net and dslreports/speedtest and the numbers are all over the place, as much as 15+ over download speeds at one test then 10 under my download speeds on the next test. Upload is way out of control too. I'm seeing 15.5mbps upload speeds. Also, dslreports keeps hanging on all of the grades. Overall, Bufferbloat, Quality, and Speed, don't show up. I've waited almost 5 minutes one time. I tried using Microsoft Edge after it didn't work on Firefox and it did the same thing. I may be in the middle of an issue with my MB8600 Cable modem. The OFDM PCL channel has an increasing number of uncorrected packets so I may have to put this all on hold till I get to the bottom of this.

 

[L&LD]

For getting a base reading of my ISP speed I turned off Adaptive QoS. Ran a few test from Speedtest.net and dslreports/speedtest and the numbers are all over the place, as much as 15+ over download speeds at one test then 10 under my download speeds on the next test. Upload is way out of control too. I'm seeing 15.5mbps upload speeds. Also, dslreports keeps hanging on all of the grades. Overall, Bufferbloat, Quality, and Speed, don't show up. I've waited almost 5 minutes one time. I tried using Microsoft Edge after it didn't work on Firefox and it did the same thing. I may be in the middle of an issue with my MB8600 Cable modem. The OFDM PCL channel has an increasing number of uncorrected packets so I may have to put this all on hold till I get to the bottom of this.

Yes, the dslreports.com/speedtest page is still not working properly.

No matter, simply click the 'Results' green badge to see the scores achieved.

Trying to get similar results from different testing sites isn't recommended. Stick to one or the other, but dslreports.com offers a much better testing and optimizing site than any others I've used.

See if the modem can be replaced or the issues otherwise corrected. That may be all you need to see better results with all your devices on a single router.

 

[randomName]

Well the people over at dslreports are saying my screen shot of my Tx connection looks fine while a tech, 2 months ago, said my Tx signal was way out of range on the high side, while the one on the phone last night said it was slightly high. I'm gonna let that go for now and just focus on the router.

I ran several test at dslreports/speedtest without QoS to get a better reading of my Rx and Tx. Rx is around 35mbps and the Tx is around 2.5mbps.

Bufferbloat grade is B

Idle = 27
Download = 45
Upload = 94


I've enabled Adaptive QoS with the settings:

Adaptive QoS
Manual Settings
fq_codel
Tx - 2.25
Rx - 29

For 'Mode' I chose customize and used the recommended priority list from FreshJR according to post#2 of his thread.

Gonna wait a few more minute and run another dslreports/speed test.

EDIT: Oh, do I need to set anything under the 'Classification' tab in the Adaptive QoS section? Particularly for PC gaming?

 

[randomName]

10+ min later and I reran the test. Bufferbloat was an A+

Idle = 23
Download = 29
Upload = 28

Are those good scores or can they be lowered?

 

[L&LD]

Well the people over at dslreports are saying my screen shot of my Tx connection looks fine while a tech, 2 months ago, said my Tx signal was way out of range on the high side, while the one on the phone last night said it was slightly high. I'm gonna let that go for now and just focus on the router.

I ran several test at dslreports/speedtest without QoS to get a better reading of my Rx and Tx. Rx is around 35mbps and the Tx is around 2.5mbps.

Bufferbloat grade is B

Idle = 27
Download = 45
Upload = 94


I've enabled Adaptive QoS with the settings:

Adaptive QoS
Manual Settings
fq_codel
Tx - 2.25
Rx - 29

For 'Mode' I chose customize and used the recommended priority list from FreshJR according to post#2 of his thread.

Gonna wait a few more minute and run another dslreports/speed test.

EDIT: Oh, do I need to set anything under the 'Classification' tab in the Adaptive QoS section? Particularly for PC gaming?

I'm confused a little with your numbers above. Without QoS, you get 'B' grade for bufferbloat and 45/94 down/up speeds?

 

[L&LD]

10+ min later and I reran the test. Bufferbloat was an A+

Idle = 23
Download = 29
Upload = 28

Are those good scores or can they be lowered?

You can't get better than an A+ for bufferbloat.

What you want to test further, if you want, is if you can keep that A+ and increase both downloads and upload limits as much as possible.

Have you tested yet with your normal use? Is the 'net a touch more responsive when you're using more than one client device at a time?

 

[randomName]

I'm confused a little with your numbers above. Without QoS, you get 'B' grade for bufferbloat and 45/94 down/up speeds?

My speeds with Adaptive Q0S Disabled = Rx ~ 35mbps, Tx ~ 2.5mbps. The buffer bloat grade was B. The buffer bloat in milliseconds for idle was 27ms, download was 45ms, upload was 94ms.

My speeds with Adaptive QoS Enabled and setup like previously mentioned = Rx ~ 28mbps, Tx ~ 1.5mbps. The buffer bloat grade is A+. The buffer bloat in milliseconds for idle was 23ms, download was 29ms, upload was 28ms.

Adaptive QoS
Manual Settings
fq_codel
Tx - 2.25
Rx - 29

 

[L&LD]

My speeds with Adaptive Q0S Disabled = Rx ~ 35mbps, Tx ~ 2.5mbps. The buffer bloat grade was B. The buffer bloat in milliseconds for idle was 27ms, download was 45ms, upload was 94ms.

My speeds with Adaptive QoS Enabled and setup like previously mentioned = Rx ~ 28mbps, Tx ~ 1.5mbps. The buffer bloat grade is A+. The buffer bloat in milliseconds for idle was 23ms, download was 29ms, upload was 28ms.

Adaptive QoS
Manual Settings
fq_codel
Tx - 2.25
Rx - 29

Okay, gotcha!

Just concentrate on keeping that A+ bufferbloat score now and see how fast you can allow your connection to go before the bufferbloat score suffers. For both downloads and uploads.