Reduced performance with AirPort Extreme

[driggins]

My LAN is currently using a Linksys WRT54GS wireless router. It's performed well, but it has some drawbacks:
-limited to 802.11G
-unable to assign a real/static IP to the router and private, virtual IP's to my LAN clients
-cannot create both secure and guest wireless networks
-non gigabit

I've purchased an Apple AirPort Extreme, which provides all of the features I want. However, when I do a simple swap of the two routers, I've noticed a significant drop in LAN and WAN performance. Weird.

Any ideas as to why the Apple router would not perform as well? Any suggestions on how to troubleshoot or fix this situation?

 

[jdabbs]

My LAN is currently using a Linksys WRT54GS wireless router.
-unable to assign a real/static IP to the router and private, virtual IP's to my LAN clients

Depending on what you mean by "virtual," I don't believe that's the router's fault.

Any ideas as to why the Apple router would not perform as well? Any suggestions on how to troubleshoot or fix this situation?

Assuming the issue is entirely configuration-related, DNS could be a culprit.
Before fiddling with settings though, it's generally a good idea to pinpoint the issue. First I would attempt to determine if the issue is throughput or something else. Try comparing Internet speed tests for the different routers. What issue are you experiencing with LAN performance?

 

[driggins]

Sorry if I don't know the proper network terminology. By "virtual" IP addresses, I am referring to assigning addresses in the 192.168.0.x range to all of my DHCP clients. Simultaneously, I want to be able to assign a static address to the router itself that is in the 216.227.x.x range (provided to me by my ISP) so that I can remote connect. The Linksys automatically uses the 216.227.x.x range for DHCP assignments. There does not appear to be the ability to override. Conversely, the Airport allows me to set one IP for the router and a different range of private address for DHCP clients.

My users reported slower performance on WAN and LAN using the Airport router.

I did some basic internet (WAN?) speed testing using speedtest.net and found:
-The ping appeared greater using the Linksys router (Airport was consistent 175 ms vs. Linksys 200-325 ms).
-Upload and Download time was basically the same for both (4 MBps)

I'd like to learn how to troubleshoot LAN performance, but that subject is entirely new to me. I am aware of Wireshark, but don't know how to interpret results. Any suggestions on how to get started?

Thanks.

 

[jdabbs]

Sorry if I don't know the proper network terminology. By "virtual" IP addresses, I am referring to assigning addresses in the 192.168.0.x range to all of my DHCP clients. Simultaneously, I want to be able to assign a static address to the router itself that is in the 216.227.x.x range (provided to me by my ISP) so that I can remote connect. The Linksys automatically uses the 216.227.x.x range for DHCP assignments. There does not appear to be the ability to override. Conversely, the Airport allows me to set one IP for the router and a different range of private address for DHCP clients.

The GS supports both WAN and LAN IP configuration. I believe you assigned the ISP-provided WAN IP to the router's LAN interface, which affected the DHCP scope.

My users reported slower performance on WAN and LAN using the Airport router.

I did some basic internet (WAN?) speed testing using speedtest.net and found:
-The ping appeared greater using the Linksys router (Airport was consistent 175 ms vs. Linksys 200-325 ms).
-Upload and Download time was basically the same for both (4 MBps)

I'd like to learn how to troubleshoot LAN performance, but that subject is entirely new to me. I am aware of Wireshark, but don't know how to interpret results. Any suggestions on how to get started?

Thanks.

I've had a user thank me for improvements in network performance before the replacement equipment left the shrinkwrap. It's important for you to identify a problem, as well as verify its existence. I imagine a user that is able to distinguish between LAN and WAN traffic is able to give you a better description other than "it's slow." At this point, talking with your user is better than running a suite of tools that may not indicate the actual problem.

The symptoms indicate the problem. Solutions are problem-specific.
On the WAN side, it doesn't sound like throughput is affected.

Troubleshooting LAN performance is straightforward. Check throughput, check link quality. Determine how widespread the problem is. Fiddle with settings or swap components until the problem goes away.


I don't recommend using Wireshark as a diagnostic. If you don't have a strong foundation already, you'll waste time worrying about issues/misconceptions like TCP checksum errors or SNMP packets consuming half your "bandwidth."