Router B set up as Access Point interfering with ping on devices connected to Router A?

[blyzr]

I have a router (Router B) set up as an AP and for WiFi only, and then a modem/router (Router A) combo from my ISP that is used for Ethernet only.

I was having terrible ping spikes intermittently, and initially fixed it with a reboot of the modem/router combo.

Then it gradually started to spike again (http://i.imgur.com/C94sWQS.png), then I decided to reboot the AP router, and immediately saw the ping drop back down ~20ms.

http://i.imgur.com/NnPCmKC.png

Could I have set up the AP router in a less than ideal fashion to cause it to interfere with ping on the modem/router? I've disabled DHCP and everything else on this guide

 

[RealBeast]

Did you set the 2.4GHz wireless channels to two different non-overlapping channels (choosing from 1, 6, and 11) on the router and AP so they do not interfere with one another or disable the radio(s) on the router?

 

[blyzr]

Did you set the 2.4GHz wireless channels to two different non-overlapping channels (choosing from 1, 6, and 11) on the router and AP so they do not interfere with one another or disable the radio(s) on the router?

I have both 2.4 and 5GHz disabled on Router A

 

[RealBeast]

You need to troubleshoot it from one end to the other then. What models are A and B?

Is the issue only on wireless, or does it also occur if you connect a computer to A with a cable?

So A and B are connected by an Ethernet cable LAN port to LAN port, have you tested the cable?

On B, DHCP is off and the device has an address that is in the network range but outside of A's DHCP assignment range?

Any wireless devices near B, like phones, baby monitors, microwave ovens?

 

[blyzr]

You need to troubleshoot it from one end to the other then. What models are A and B?

Is the issue only on wireless, or does it also occur if you connect a computer to A with a cable?

So A and B are connected by an Ethernet cable LAN port to LAN port, have you tested the cable?

On B, DHCP is off and the device has an address that is in the network range but outside of A's DHCP assignment range?

Any wireless devices near B, like phones, baby monitors, microwave ovens?

I ended up reverting my Nighthawk (Router B), to stock firmware, I had dd-wrt firmware installed. It seems to be working with the much simpler AP setup on Netgears firmware, I must've missed something on the dd-wrt AP setup which was causing conflicts somewhere. Maybe IP conflicts since Router A's IP range is ...168.2-...168.254, so not sure if there was any wiggle room for an AP's IP without conflict.

Seems to be all good now though!

 

[RealBeast]

The access point should be in the same network range as the router, so say 192.168.1.1 for the main router and 192.168.1.254 for the AP.

If they are in different subnets then you need to attach the AP with its WAN port and designate the primary router address as the gateway WAN for the AP and let the AP assign DHCP addresses to devices connected, which is more complex than needed for home networks.

 

[blyzr]

The access point should be in the same network range as the router, so say 192.168.1.1 for the main router and 192.168.1.254 for the AP.

If they are in different subnets then you need to attach the AP with its WAN port and designate the primary router address as the gateway WAN for the AP and let the AP assign DHCP addresses to devices connected, which is more complex than needed for home networks.

Wait so I shouldn't plug in to the WAN port as of now? It's plugged into the WAN could that be causing the issues?

 

[RealBeast]

An AP is connected LAN to LAN and has an address in the network range.

 

[blyzr]

An AP is connected LAN to LAN and has an address in the network range.

So I switched it from the WAN to LAN, but still getting ping spikes, and then the ping is fine when restarting or turning off the AP router.

Any ideas what could be the issue with the AP router causing ping spikes into the high 100s when it's on?

 

[blyzr]

You need to troubleshoot it from one end to the other then. What models are A and B?
A is a SuperHub from Virgin Media, Router/Modem combo. B is a Netgear Nighthawk R7000

Is the issue only on wireless, or does it also occur if you connect a computer to A with a cable?
It is only an issue on Ethernet connected to Router A, devices connected to the AP (Router B) aren't really using any programs affected by ping spikes

So A and B are connected by an Ethernet cable LAN port to LAN port, have you tested the cable?
The cable was being used before integrating the AP and had no issue.

On B, DHCP is off and the device has an address that is in the network range but outside of A's DHCP assignment range?
Not sure on DHCP assignment range as the Router A settings are quite primitive. http://i.imgur.com/paOInDp.png

Is the "IP Address" the DHCP assignment range? If so should I maybe put it to 192.168.0.3, and then set the AP to 192.168.0.2?

Any wireless devices near B, like phones, baby monitors, microwave ovens?
Nothing near it other than Router A

 

[RealBeast]

I would reset the AP router and set it up again from the beginning.

LAN to LAN connection, insure that the cable connecting to the router is good, DHCP off on the AP and set it to an address in the network range that is not in the DHCP assignment range of the main router (so that a second device is not assigned that address).

Then insure that there are no sources of interference near the AP.

Run a series of constant pings through the AP to your main router address for a few minutes to test your latency. It is not valid to test pings outside the network as that could really be a problem beyond your main router -- i.e. with your ISP.

So open a command prompt window and type:

ping -t 192.168.0.1 (or whatever the address of your main router is), hit enter then allow it to run a number of pings, then hit ctrl-c to stop and look at the results.

 

[RealBeast]

Okay, saw your latest post after my most recent.

Yes, change the DHCP range of A to 192.168.0.3 and up and then make the AP 192.168.0.2.

Run that ping -t test that I suggested on a wired device and use a common address like 8.8.8.8 (Google DNS) for the address and see if you note latency spikes. If so, then run a tracert command (tracert 8.8.8.8) to see where the spike is occurring to insure that it is inside your network as opposed to at your ISP level.

edit: also attach one of the wired devices to an LAN port on the AP and run some pings to see if that makes a difference.

 

[blyzr]

Okay, saw your latest post after my most recent.

Yes, change the DHCP range of A to 192.168.0.3 and up and then make the AP 192.168.0.2.

Run that ping -t test that I suggested on a wired device and use a common address like 8.8.8.8 (Google DNS) for the address and see if you note latency spikes. If so, then run a tracert command (tracert 8.8.8.8) to see where the spike is occurring to insure that it is inside your network as opposed to at your ISP level.

edit: also attach one of the wired devices to an LAN port on the AP and run some pings to see if that makes a difference.

Ping from computer on the AP to main router is <1ms.

Ping seems to have stabilized somewhat after setting AP IP outside of DHCP range of Router A (still spiking to 100ms at times.) http://i.imgur.com/WGxSgiJ.png (Big red packet loss is where I reset both routers and set new IP on the AP)

In the past the hops that have had the highest ping are the two virginmedia.net hops, but will continue to monitor to see if setting IP outside of DHCP range has fixed the issue.

Thanks for all your help!

edit: I ran some tests whilst connected to the AP via ethernet and had similar fluctuations in ping, going up to ~90ms for a couple pings, but mostly around 15-30ms.

 

[blyzr]

Quick question. Could someone on the network be affecting the ping so drastically? Reason I ask is after spiking to 100ms pretty constantly I noticed it's normalized around 20ms after one of the computers was shutdown on the network. I'll take a look at that PC in the morning, but anything I should look for in particular?

 

[blyzr]

I just had a look at the logs to try and match when the spiking stopped.

Not sure what to make of this http://i.imgur.com/vqRyXCq.png

 

[RealBeast]

Quick question. Could someone on the network be affecting the ping so drastically? Reason I ask is after spiking to 100ms pretty constantly I noticed it's normalized around 20ms after one of the computers was shutdown on the network. I'll take a look at that PC in the morning, but anything I should look for in particular?

Sure, a user with many connections or those that are upload intensive (like a torrent service for example) would adversely impact the other users.

 

[RealBeast]

I may misunderstand the testing that you are doing.

The latency in a wired network should always be around 1ms or less from any point to any other on that network, not using a site outside the network since that introduces an unknown variable that is not important to the local network diagnostic. I would look at all cables, adapters, and network devices (routers/switches/etc) to determine where the delay is coming from most likely due to high packet loss. If one machine is running a high connection bandwidth intensive program like a torrent server you may experience delays but they should not show up in the point to point testing of each segment.

And in the case of wireless, latency is highly dependent upon the amount of wireless bandwidth being used, although I understand that you are using an all wired network.