Jump to content
RESET Forums (homeservershow.com)

Router and DNS Latency


awraynor
 Share

Recommended Posts

Moved prior discussion out of the Amplifi thread. 

 

 

@itgeeks

 

 

ping main Eero:  Minimum = 0ms, Maximum = 0ms, Average = 0ms

 

ping 2nd Eero: Minimum = 3ms, Maximum = 6ms, Average = 4ms

Another run it gave me: Minimum = 23ms, Maximum = 97ms, Average = 62ms

 

It's only about 15 feet away from main unit and one wall in between.

 

ping 3rd Eero: Minimum = 2ms, Maximum = 6ms, Average = 3ms

 

google.com:  Minimum = 13ms, Maximum = 43ms, Average = 25ms

 

yahoo.com: Minimum = 44ms, Maximum = 73ms, Average = 54ms

 

My Eero is set to use Charter's DNS. I've changed it to Google for DNS

and can't say I felt it was any different.

 

 

So with the above information is there any likely improvement doing MoCA to each Eero. Response seems pretty good. 

What do you use for DNS? External times seem pretty high?

Link to comment
Share on other sites

How are you running these comparisons? You have a laptop and connecting to specific modules? Command line ping?

Link to comment
Share on other sites

I'm on uverse and used to see all sorts of weird delays loading pages (the symptoms would come and go though). I ended up pointing my router at googles dns servers and that mostly resolved my issues. AT&T must not be able to serve DNS very reliably.

 

To test latency through your eero, I would suggest doing ping tests to the first hop into charter, so you are trying going through eero, but not reliant on anything on the "internet". Try to connect your test device via each eero device and test the same destination. You should get pretty consistent results for each test and I would hope you would only see a few ms variance between tests.

 

 

Sent from my iPhone using Tapatalk

Link to comment
Share on other sites

Command line ping from desktop that is hardwired to switch that is connected to main Eero. 

 

Listening to averageguy podcast at the moment so don't want to change to Google DNS at this second.

 

I did the other day and didn't seem to make any difference in real world use. 

Link to comment
Share on other sites

Setting your DNS server to Google won't affect your ping times.  It might help with apparent latency when loading webpages, though ...

 

As a data point for you, I ran "tracert -d 8.8.8.8" from my windows desktop and found the first public IP, which would be the first hop into uverse for me.  Doing a ping test of 10 pings to that IP, I see a min/avg/max of 20/20/22ms.  Doing another ping test to 8.8.8.8 results in 24/24/26ms.  So, AT&T is a base 20ms away from me and Google is another 4ms past them.

 

The fact that you are seeing ping results of 13/25/43ms to google tells me that your connection is very inconsistent, so the trick is finding which hop is introducing all of the variance.

 

Personally, I would take the tracert results and run tests to each hop ... trying to find which hop no longer produces consistent results.  Perhaps Charter happens to have a routing problem and they are routing traffic in weird ways so each ping test gets send down different paths, resulting in huge ping times.  Maybe it's outside of Charter's network ... hopefully it's not inside your own home.

Link to comment
Share on other sites

I shut down every PC in the house except for the test PC to check those times. 

 

I ran netalyzr as referenced by Drashna. The results are a little concerning?

 

3 popular names have a moderate anomaly: we are unable to find a reverse name associated with the IP address provided by your ISP's DNS server, although we expected to find a name. This is most likely due to a slow responding DNS server. If you rerun Netalyzr and see this condition remain, it could be due to a misconfiguration on the part of the domain owner, deliberate blocking using DNS, or your DNS server could be misconfigured or enabling a Man-in-the-Middle attack.

 

Network performance (?): Latency: 65 ms, Loss: 0.0% +

 
TCP connection setup latency (?): 40ms +
The time it takes your computer to set up a TCP connection with our server is 40 ms, which is good.
 
TCP connection setup latency (?): 40ms +

 
Background measurement of network health (?): no transient outages +
During most of Netalyzr's execution, the client continuously measures the state of the network in the background, looking for short outages. During testing, the client observed no such outages.
 
Background measurement of network health (?): no transient outages +

 
Network bandwidth (?): Upload 590 Kbit/s, Download >20 Mbit/s +
Your Uplink: We measured your uplink's sending bandwidth at 590 Kbit/s. This level of bandwidth works well for many users.
During this test, the client observed 2 reordered packets.
Your Downlink: We measured your downlink's receiving bandwidth at >20 Mbit/s. This level of bandwidth works well for many users.
During this test, the client observed 58 reordered packets.
 
Network bandwidth (?): Upload 590 Kbit/s, Download >20 Mbit/s +

 
 
Network buffer measurements (?): Uplink 900 ms, Downlink 130 ms
We estimate your uplink as having 900 ms of buffering. This level can in some situations prove somewhat high, and you may experience degraded performance when performing interactive tasks such as web-surfing while simultaneously conducting large uploads. Real-time applications, such as games or audio chat, may also work poorly when conducting large uploads at the same time.
We estimate your downlink as having 130 ms of buffering. This level may serve well for maximizing speed while minimizing the impact of large transfers on other traffic. 
Link to comment
Share on other sites

Oh, and to comment on your 2nd eero test where you saw ping times of 23/62/97ms ... that's normal from what I can tell ... at least, that's the kind of thing I see when I ping from my wired desktop to a wireless macbook ... depending on the wifi interference, if my mac is doing "other things" at the same time, etc... I see anywhere from 1ms to 104ms ping times, and that's across 5Ghz 802.11ac where I am getting about 200Mbps iperf3 results.  Interestingly, when I started the iperf3 test, pings from the desktop consistently showed 7-8ms, then when iperf3 finished, it jumps around a bit, hitting 1228ms once.  So, yes, wireless ping times can be *very* inconsistent in general, it seems.  During all of my iperf3 tests, it ran 317 pings from my desktop and I saw 0/41/1228ms for min/avg/max. :)

Link to comment
Share on other sites

Of course, doing ping tests from my macbook that's on wireless, it generally produces much more consistent results... pinging google's DNS servers I see:

 

on 2.4Ghz wireless:

--- 8.8.8.8 ping statistics ---
30 packets transmitted, 30 packets received, 0.0% packet loss
round-trip min/avg/max/stddev = 25.654/28.610/39.420/2.941 ms

 

on 5Ghz wireless:

--- 8.8.8.8 ping statistics ---
30 packets transmitted, 30 packets received, 0.0% packet loss
round-trip min/avg/max/stddev = 26.226/27.682/29.747/0.908 ms

 

I then did the same ping test, but then also looked at my detailed wireless device connection stats, so it's probing for RSSI values, etc... and got:

 

--- 8.8.8.8 ping statistics ---
30 packets transmitted, 30 packets received, 0.0% packet loss
round-trip min/avg/max/stddev = 25.787/62.761/248.472/52.241 ms

 

So yeah ... testing ping latency over wireless can be very interesting ...

Link to comment
Share on other sites

Going to try to switch DNS services and see if that helps in lieu of buying a bunch of new MoCA or wireless equipment. 

 

Been with Charter for years and have had similar issues. My only other  option if capped AT&T service which isn't an option. 

Link to comment
Share on other sites

Thanks for prompting me to learn a tiny bit about MoCA adapters... ;)

 

The MoCA adapters should give you much more consistent latency between your eero nodes, if you decide to try that out as a way to use the ethernet backhaul.  Poking around it looks like MoCA 1.1 devices average latency of around 4.5ms (supposedly with a max latency of 7.5ms) and MoCA 2.0 drops it to 3.6ms.  MoCA 1.1 also seems to have throughput of around 175 Mbps (275 Mbit/s PHY rate), MoCA 2.0 bumps that to 400 (basic) or 800 (advanced) using 700 Mbit/s and 1.4 Gbit/s PHY rates, and MoCA 2.5 supposedly allows up to 2.5Gbps (https://en.wikipedia.org/wiki/Multimedia_over_Coax_Alliance#Technology).

 

Unfortunately, from what I can tell, MoCA 2.5 devices don't exist yet.  So, while you might stabilize the latency a bit, you might be artificially limiting your throughput ... I don't know what kind of throughput the eero nodes can actually pump over wireless or if it's over 175 Mbps.  Looks like amazon sells the actiontec MoCA 2.0 for $141 for a 2-pack ... that's a lot to add to the already expensive costs of eero, IMO (cause you'd need at least 2 of these).

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
 Share

×
×
  • Create New...