Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[trex76]

I recently purchased a Cisco IP Phone that I added to my network.  I decided for maximum reliability I would connect it to a switch I had in place connected to my DIR 655 router with FW 2.10NA.  While I was running the cable, I ran two additional cables for my primary computer(MacBook Pro) and a macmini.  Everything worked fine initially, but after a few hours(probably 5) everything slowed to a crawl and only some sites worked (google and yahoo) while others didn't.  It appeared to be a DNS issue.  Eventually no sites worked.  Nothing fixed the issue until I power cycled the modem and then the router.  Then everything came back up, again a few hours later Internet issues started occurring.  I did a bunch of troubleshooting, looking at articles on this forum and other sites and nothing has resolved the problem.  Prior to connecting the phone, macbook, and macmini I had 4 other devices connected to the D-Link DGS-2208 switch.  The devices connected were a PS3, two DirecTV DVRs, and a Apple Airport Express used for sending Music to a Stereo Receiver.  I have since put everything back to my original configuration and connected the Phone via the wireless network as well and everything appears to be running smoothly.  I would prefer to use the switch for LAN traffic, but it appears when I do this I overload the router and that isn't optimal.  Has anyone seen similar behavior when using a switch connected to the DIR-655?  Is there any settings I should look at to address this problem or am I looking at purchasing a bigger router if I want to utilize my switch at full capacity?

Thanks

[FurryNutz]

Preferred method is to connect the switch to the the 655 and all other wired lan devices to the switch. However, for devices that need a more direct connection to the Internet via the router, it's recommended to connect those devices directly to the back of the DIR-655. For example, I have to gaming consoles, they are both directly connected to the back of the router for better management by the router and if other wired devices are communication between the switch, they won't be interfered by the devices directly connected to the router.

You might want to put the IP phone into the DMZ on the 655 as well.

Here how my setup looks like:
ISP Modem><Router><Game consoles connected to back of router
                           ><24Pt Gb switch connected to back of router><All other wired devices connected to the switch.

I can watch a movie from my file server on my boxee box via the switch while gaming is going on using the router.

[trex76]

I would agree with you on the preferred method.  I would like to have everything connected to the switch.  I don't think I need the phone connected directly to the router as it works fine as is.  What I don't get is why when I hook up 3 things to the switch, two of which were previously connected via wireless, the router is getting bogged down and eventually not working?  I assume it has to do with the Gigabit vs. Wireless connection or the switch is doing something funny, but I'm not sure whether it's the memory on the router or DNS related or something completely different.  It just seems strange to me that changing 3 devices from Wireless to Wired connections bring the router to its knees.  Thanks for your response.

[FurryNutz]

Only means that something is either not correctly connected, configured or something is bandwidth hoging the connections to the point that the router is slowing down. I would turn OFF everything and disconnect the switch for now. Connect up one PC first, then test some internet or web sites. Turn on another PC or device, keep doing this to see if the router starts to slow down.

[trex76]

I suspect you are right.  In fact, I have a feeling this has been happening even when I have my devices connected via wireless, it was just taking a lot longer.  I'll be checking devices one by one over the next few days.  One thing I noticed is when I ping the router from the LAN I get every other ping with bad response times.  Here is an example:

PING 192.168.15.1 (192.168.15.1): 56 data bytes
64 bytes from 192.168.15.1: icmp_seq=0 ttl=64 time=1.612 ms
64 bytes from 192.168.15.1: icmp_seq=1 ttl=64 time=35.790 ms
64 bytes from 192.168.15.1: icmp_seq=2 ttl=64 time=1.909 ms
64 bytes from 192.168.15.1: icmp_seq=3 ttl=64 time=35.271 ms
64 bytes from 192.168.15.1: icmp_seq=4 ttl=64 time=2.118 ms
64 bytes from 192.168.15.1: icmp_seq=5 ttl=64 time=35.763 ms
64 bytes from 192.168.15.1: icmp_seq=6 ttl=64 time=4.031 ms

When things aren't working, it's more like 200 ms on the low end and 500 ms on the high end.  This may very well be that with the wireless connection it takes longer for the offending device to flood the network.  With the wired connection it happens much faster since it has a gigabit connection to the switch and the switch has a gigabit connection to the router.

[FurryNutz]

Start with 1 PC and go from there. Disconnect everything. Other connected devices can and are probably causing some additional router processing which could be the cause of the odd ms times.

Keep us posted on your results.

[trex76]

I think I may have found the culprit.  I've only been running clean for an hour so I'm a little on the optimistic side, but this seems to have a good chance of working.  For my job I have VMWare loaded on my MacBookPro running Windows 7 .  I started up wireshark on the MacBook and noticed a ton of NetBios traffic destined for 192.168.100.82.  This seemed strange to me since my private network is on a totally different 192.168 network and nothing should be configured with that IP.  I went into the TCP/IP settings on my Windows 7 Virtual Machine and disabled NetBios.  Restarted Wireshark and all the NetBios traffic was gone.  After getting that resolved I did some more exploration to see where that address was coming from.  I connect to a VPN for work and the address is in the 10.0.0.0 reserved space, but it gets a DNS address of 192.168.100.83.  So my best guess is my local adapter was using the NetBios info from the VPN connection and sending it to it's default gateway(MacBook Interface) which eventually got to the Router.  The router had no clue where 192.168.100 traffic should go so it sent it to the ISP which promptly dropped the packets.  It may have even been causing violations on the ISP so they blocked me.  I've seen that with other ISPs, if you send traffic destined for a private address your connection gets reset.   I'll repost in a day or so if my network is still running clean, but I'd say there is high probability the NetBios traffic was causing problems.

Thanks for the assistance

[FurryNutz]

Glad you found something. Hope this was the root cause and this will fix it.

Keep us posted.

[trex76]

I have not had any network issues since I disabled NetBios on the Local Network Adapter on my Windows 7 Virtual Machine.  Ping times between my MacBook and the router have been less than 1 ms since that time.  I've hooked everything back up to the switch and I'm still running clean.  Thank goodness for wireshark, I don't think would have seen the NetBios traffic otherwise.  Problem solved, no need to buy a new switch or router.

[FurryNutz]

Glad you found the fix.

Enjoy.