Slow powerline with green link quality light

[sliverstorm]

My powerline adapters (The Actiontec 500 plugs) report via the status LED that they have excellent signal- according to the Actiontec web page, green means "100Mbps+".

Yet when I test transfers to my NAS, I get 4MBps where I had 20MBps+ before over wireless N450.

Now, I understand that a powerline adapter probably can't match N450 at close range, but 4MBps (32Mbps) seems much too low given the green light. Is this normal? As I *know* my NAS performs better over wireless, all I can think is it is either the powerline adapters, or perhaps my RT-N66U is giving me terrible switch performance (NAS and powerline adapter both connected to RT-N66U LAN ports)

Any feedback? Thanks!

 

[thiggins]

Look for noise sources.
http://www.smallnetbuilder.com/lanw...ug-five-ways-to-boost-powerline-network-speed

 

[sliverstorm]

Hmm... I've disconnected every non-appliance that I can in my house (where appliance is fans, stove, heaters, garage opener, fridge, etc) and nada. At this point the only non-appliances are my computer & related devices, and my modem/router/nas. Both groups are isolated by a power strip, and I have experimented further by adding an 18ft extension cord in between the power strip and the wall.

At this point all I can think is perhaps the demand controller is introducing noise. I can't unplug that, of course. It's also from the 80's, so I am crossing my fingers it isn't generating any 2GHz+ noise.

Lights are also all off.

 

[thiggins]

Any cellphone chargers or other wall-warts?

 

[sliverstorm]

Any cellphone chargers or other wall-warts?

I did have a couple wall warts in the form of a cellphone charger, a smartradio power plug, a water fountain pump and a ham radio charger, but those are all disconnected now.

I have been periodically walking my house to see if there's anything I forgot.

 

[thiggins]

I have found cellphone chargers are the worst offenders. They can kill throughput even when in an outlet across the room.

I suggest you get the two adapters in the same room and run a throughput test there. That will tell you whether you what your maximum possible throughput is.

I never pay attention to the status lights on these things.

Download the TRENDnet powerline utility from here and see what it gives you for link rates. Better than trying to figure out what is going on from looking at the lights.

 

[sliverstorm]

Yup, right now the two adapters are in the same room, opposite walls.

Should the TRENDnet utility detect Actiontec adapters? It isn't seeing them, even when I reset the key on the adapters and try to copy the default from the sticker into the TRENDnet utility.

 

[thiggins]

It should. You need to run the utility on a computer directly plugged into one of the powerline adapters.

 

[sliverstorm]

Ok, I got the utility sorted out, I had an old'n'busted installation of WinPCap. Installed a new version, it finds the devices now.

It seems to be pretty happy:

 

[thiggins]

Looks like that is max link rate. Are you still getting 4 MB/s transfers?

 

[CaptainSTX]

Powerline Monitoring

It should. You need to run the utility on a computer directly plugged into one of the powerline adapters.

That is different than how my Netgear XET1001 Powerline Encryption utility works. I can monitor the link speed of any adapters attached to my network using a PC I have the utility installed on even if I am connected to my network/router using WiFi as long as I am in the same subnet.

 

[sliverstorm]

Looks like that is max link rate.

That's what's so confusing!

Are you still getting 4 MB/s transfers?

Yes

 

[thiggins]

How are you measuring transfer rate?

 

[sliverstorm]

How are you measuring transfer rate?

2GB+ files copied from my NAS to my desktop's SSD using Explorer.

To double-check, I hopped back on the wireless just now and got 11MBps (slower than it has been in the past, but still clearly faster)

Also just for kicks I tried FTP'ing from the NAS w/ Firefox to take Explorer and SMB out of the equation- it was even slower

P.S. Why am I doing this if wireless is faster? My wireless has worse ping & drops packets, and the reception in other rooms is not very good- but I am attempting to prove out the powerline in the same room before using powerline between rooms.

 

[thiggins]

Are you copying a folder or one large file? If a folder, try one large file.
Make sure you are copying in the same direction.
Also try the opposite direction
Make sure wireless is DISABLED when you are copying via powerline

 

[sliverstorm]

Are you copying a folder or one large file? If a folder, try one large file.

One large file

Make sure you are copying in the same direction.

Yup, always pulling from the NAS

Also try the opposite direction

Ok, this is pretty wild, copying TO the NAS (a DS213+) is 11MB/s+, which is about as good as I can hope to get. These adapters have 100Mbps ports, so the theoretical maximum is 12.5MB/s

Still 4MB/s copying FROM the NAS though, which is frustrating- read is more important than write to me. Can I just switch these speeds!?

Make sure wireless is DISABLED when you are copying via powerline

Do you mean on my computer or on my router? I tried shutting off the wireless antennas on the router because it is attached to the wall and I thought it might be sending some spectrum noise into the power lines, but that didn't help. As for on my computer- it doesn't have a WiFi adapter, I use a EA-N66 (https://www.asus.com/Networking/EAN66/) which is a LAN-based wireless adapter

 

[thiggins]

Disabling wireless: I thought your computer might have wireless built in and didn't want the computer trying both (or the other) path when you were trying for powerline.

I suspect the difference is how the wireless and powerline connections are breaking up the transfers. You would have to stick Wireshark on the line and look at the actual packets to confirm.

The WiFi guys have done a lot of work to make the most of the airtime they get. They really work to minimize overhead and push as much data into a packet as possible. Probably powerline drivers aren't as smart / efficient.

Non symmetrical throughput has driven me crazy trying to figure out why. I see it in wireless and also powerline. You might be able to improve it by playing with Windows networking parameters.

But a simpler approach might be just to try another pair of powerline adapters. I'd suggest 200 Mbps adapters vs. 500 Mbps. In my experience, 200 Mbps adapters provide much more consistent performance across locations. 500 Mbps adapter throughput tends to fall off pretty quick.

 

[sliverstorm]

Non symmetrical throughput has driven me crazy trying to figure out why. I see it in wireless and also powerline. You might be able to improve it by playing with Windows networking parameters.

I'm not very hopeful, as both wireless and powerline tests have used my PC's LAN port (as the wireless adapter is LAN-based). But, what might I go about tweaking?

Edit: I suppose one key difference would be the negotiated ethernet link speed. The Actiontec will negotiate 100MB/s, while the EA-N66 negotiates 1Gb/s

But a simpler approach might be just to try another pair of powerline adapters. I'd suggest 200 Mbps adapters vs. 500 Mbps. In my experience, 200 Mbps adapters provide much more consistent performance across locations. 500 Mbps adapter throughput tends to fall off pretty quick.

Ok, I've put in an order for some 200Mbps units. I was hoping to take advantage of the greater ultimate bandwidth of the 500Mbps units even if per-unit link speeds were a bit lower, but this is just too much worse.

 

[sliverstorm]

Well, this is disheartening. Same performance out of the 200Mbps TP-Link adapters, although it's slightly greater than 4MB/s rather than slightly less.

 

[thiggins]

Well, this is disheartening. Same performance out of the 200Mbps TP-Link adapters, although it's slightly greater than 4MB/s rather than slightly less.

at least now you know the problem is not in the adapters.