High Ping Time on Local Network

Since my other articles on figuring out problems that you can’t Google are popular I thought I would write about my latest mood killer. I began to get a very high ping time on my local network, even higher out to the Internet. This caused problems with ssh sessions and NVIDIA game streaming, but was not noticeable when watching Netflix.

It started simply enough, one night I left my wife to watch Netflix downstairs while I went up stairs to get some work done. My job involves using ssh to login to remote servers, any latency (high ping) is very noticeable. That was the case this night. I could type much faster than the text would display, using the arrow keys to navigate was jerky at best. I figured my ISP was being lame. I finished up as quickly as possible. The next day there was no problem.

Jump cut to a few days later. Same situation. I head upstairs to work and my ssh session is just crap. The next day its fine.

The next morning is a Saturday. I take a seat on the couch and try to play Grim Dawn over NVIDIA game streaming and it doesn’t work. The bad connection icon appears in the top right, the graphics are pixelated and jerky, not good. Having no clue what could be wrong, I give my PC and Shield TV a restart. No luck, same problem. I tried the test connection option on the Shield TV, it reported back that I have a 9mb/s connection and 200ms ping. Considering I was on a wired gigabit network with only a switch between the Shield TV and my PC this was very bad.

My wife and children, bored of me, starting watching Netflix as I furiously Googled for a hint at a solution. Immediately my phone’s connection started having problems. Sites wouldn’t load or would load slowly. I tried a speed test and while my bandwidth was ok, my ping was 500ms. I tried pinging my PC, the ping times were all over the place, 1ms, 50ms, 100ms, and everything in between. Netflix worked fine, which I chalked up to streaming not being interactive and thus not caring about latency.

I at this point I knew it was something with my local network affecting both wired and wireless, but I had no idea what.

I had several command prompts open pinging devices around my house. All showed high pings. I expected less than one millisecond and the lowest was 5ms and very jittery.

I then thought about the problems I had with UNRAID. What if another port on my router’s switch is bad? My wife stopped watching Netflix and went into another room. I unplugged everything from my router and tried wifi, all good. I started plugging in all of my devices one at a time, testing in between each one. All good again. I opened up a couple of command prompts and started continuous pings, all good. I tried the NVIDIA game streaming test, the test failed just as before. I noticed something new this time though, my continuous pings were fine until the streaming test; during the test they were high. As soon as the test stopped they went back to normal. I started a movie on Netflix, the pings went bad again. I moved the Shield TV to a different port on my switch. The game streaming test succeeded. Netflix could play without jacking up the pings. It was another bad port in the router’s switch.

That left me with a switch that had four gigabit ports. One would say it was connected at a gigabit, but only pass 100mb/s of traffic. Another port messed up my entire wired and wireless network if any decent amount of traffic was passed. As a quickie fix I daisy chained another gigabit switch to one of the good ports. For a real long term fix I purchased another router.

Netflix was red herring. Watching Netflix would pass traffic over the bad port in the switch, jacking up the network. When I watch Netflix I’m not doing anything else on the Internet and therefor would not see the problem.

The router causing all of this was an ASUS NT-N66U. I got four years of use out of it before sending it to the trash. In the first year I had a pretty bad problem with it dropping my WAN connection. Otherwise I was happy with it. Its a shame it didn’t last longer.

ASUS RT-N66U: Other odd system log messages explained

Aside from the, ‘WAN Connection: Ethernet link down.’, message I was seeing in my ASUS RT-N66U system log, there are two other error messages that appear often.

miniupnpd: unsupported NAT-PMP version : 2

This message appears in my log every hour on the hour. As near as I can tell it doesn’t cause any problems with my Internet connection or general network performance. I’ve figured out it only occurs when my wife’s Mac Mini is turned on. The Mac works just fine  during and outside times the error message is being logged. Either Mac OS itself or some program she has is using UPnP in a way the ASUS router doesn’t care for. I don’t have a good answer on this outside of saying it doesn’t seem to affect much.

kernel: eth1: received packet with  own address as source address

I have two access points in my home, both with the same SSID. The first access point is the one built into the ASUS RT-N66U, the second access point is an old Netgear router with everything but the access point functionality turned off.

This error occurs when a client moves from the ASUS RT-N66U to the Netgear access point. The network is general is not affected. The client moving between the two access points will lose some packets. The packet loss isn’t noticeable unless you are looking for it.

 

ASUS RT-N66U – WAN Connection: Ethernet link down.

A few weeks ago I began to have a recurring and incredibly frustrating problem with my Internet connection. Multiple times a day my Internet connection would drop for a couple of minutes at a time. My cable modem would show I had a good link. Upon checking my ASUS RT-N66U router I would see this in the system log:

WAN Connection: Ethernet link down.

Then a few minutes later the connection would come back on its own. I would see this in the log:

WAN Connection: Ethernet link up.

I work from home, mostly using SSH to log into remote servers. My connection would flake out just long enough for my SSH connection to drop. It drove me up the wall.

I did a lot of Google searching and couldn’t find much of anything about the problem. What I did find wasn’t productive.

Going strictly off of the error message I figured I had one of three problems, bad network cable between the router and modem, bad WAN port on the router, or bad WAN port on the modem. I changed network cables with no luck. I replaced my ASUS router with an old Netgear router I had. The problem stopped. I went back to the old cable, still no problem. I put the ASUS router back into play, the problem started again. I knew it had to be something with the ASUS router. I had owned it for too long to send it back for warranty coverage. I also thought it odd that the hardware in the WAN port could be flakey enough to cause an intermittent problem like this while otherwise working fine. I figured it was software related.

In order to simplify things I turned off every feature of the router I didn’t %100 need. Guest wifi network, QoS,  USB support, DDNS, cloud support, and so on and so forth were turned off. I even shut down the second access point I have in my house. After simplifying everything something wonderful happened. The problem stopped. I gave it a few days to be certain and sure enough, no more drops.

I then began to turn on each service again, one per day. When I turned on DDNS via no-ip.com the drops started again within a few hours. I turned off DDNS and the drops stopped. The culprit had been found.

I don’t have a good idea why the DDNS client causes the WAN Ethernet port to drop out. Its some sort of bug in the 3.0.0.4.376_1071 firmware I’m running.

In the end my solution was to leave DDNS turned off in the router and then to install a no-ip.com update client on my computer.

TL;DR

My ASUS RT-N66U router started dropping my Internet connection. It had something to do with updating no-ip.com. Turning off the DDNS client solved the problem.

Update 9/10/14: I switch to the Merlin Firmware and even with DDNS turned on the router has been fine with no drops. That aside the new firmware is working great, all of the features I would expect and I have not noticed any change in my speed.

Update 5/3/16: I’m still running the Merlin Firmware, updating it as they become available. No problems at all.