What causes periodically high latency on a network?

0

I have a small home network with a Hitron modem/router (model number CGN3ACSMR, software version 4.5.8.22).

Since having the router installed a month ago, I've been experiencing some odd latency patterns, both when accessing hosts outside my LAN as well as when trying to access hosts on my LAN. It's not usually intrusive, but it makes SSH annoying, and it prevents me from hosting certain services (games, streaming API servers, etc.) on my network.

Here's a dump of ping to google.com, done from my laptop, connected over the modem/router's 5GHz network:

PING google.com (209.148.198.238): 56 data bytes

64 bytes from 209.148.198.238: icmp_seq=0 ttl=57 time=18.084 ms
64 bytes from 209.148.198.238: icmp_seq=1 ttl=57 time=30.351 ms
64 bytes from 209.148.198.238: icmp_seq=2 ttl=57 time=26.911 ms
64 bytes from 209.148.198.238: icmp_seq=3 ttl=57 time=26.344 ms
64 bytes from 209.148.198.238: icmp_seq=4 ttl=57 time=149.377 ms
64 bytes from 209.148.198.238: icmp_seq=5 ttl=57 time=143.671 ms
64 bytes from 209.148.198.238: icmp_seq=6 ttl=57 time=122.085 ms
64 bytes from 209.148.198.238: icmp_seq=7 ttl=57 time=20.993 ms
64 bytes from 209.148.198.238: icmp_seq=8 ttl=57 time=92.836 ms
64 bytes from 209.148.198.238: icmp_seq=9 ttl=57 time=22.411 ms
64 bytes from 209.148.198.238: icmp_seq=10 ttl=57 time=28.901 ms
64 bytes from 209.148.198.238: icmp_seq=11 ttl=57 time=24.592 ms
64 bytes from 209.148.198.238: icmp_seq=12 ttl=57 time=31.203 ms
64 bytes from 209.148.198.238: icmp_seq=13 ttl=57 time=17.344 ms
64 bytes from 209.148.198.238: icmp_seq=14 ttl=57 time=155.770 ms
64 bytes from 209.148.198.238: icmp_seq=15 ttl=57 time=133.970 ms
64 bytes from 209.148.198.238: icmp_seq=16 ttl=57 time=22.078 ms
64 bytes from 209.148.198.238: icmp_seq=17 ttl=57 time=27.406 ms
64 bytes from 209.148.198.238: icmp_seq=18 ttl=57 time=19.005 ms
64 bytes from 209.148.198.238: icmp_seq=19 ttl=57 time=26.037 ms

--- google.com ping statistics ---
20 packets transmitted, 20 packets received, 0.0% packet loss
round-trip min/avg/max/stddev = 17.344/56.968/155.770/51.178 ms

Notice the 100+ms spikes showing up periodically.

Here are the (very similar) results of pinging an address on my LAN, once again done from my laptop:

PING 192.168.0.20 (192.168.0.20): 56 data bytes

64 bytes from 192.168.0.20: icmp_seq=0 ttl=64 time=1.015 ms
64 bytes from 192.168.0.20: icmp_seq=1 ttl=64 time=4.472 ms
64 bytes from 192.168.0.20: icmp_seq=2 ttl=64 time=1.415 ms
64 bytes from 192.168.0.20: icmp_seq=3 ttl=64 time=4.467 ms
64 bytes from 192.168.0.20: icmp_seq=4 ttl=64 time=34.398 ms
64 bytes from 192.168.0.20: icmp_seq=5 ttl=64 time=74.872 ms
64 bytes from 192.168.0.20: icmp_seq=6 ttl=64 time=54.049 ms
64 bytes from 192.168.0.20: icmp_seq=7 ttl=64 time=4.670 ms
64 bytes from 192.168.0.20: icmp_seq=8 ttl=64 time=4.442 ms
64 bytes from 192.168.0.20: icmp_seq=9 ttl=64 time=4.868 ms
64 bytes from 192.168.0.20: icmp_seq=10 ttl=64 time=0.982 ms
64 bytes from 192.168.0.20: icmp_seq=11 ttl=64 time=1.116 ms
64 bytes from 192.168.0.20: icmp_seq=12 ttl=64 time=1.645 ms
64 bytes from 192.168.0.20: icmp_seq=13 ttl=64 time=0.888 ms
64 bytes from 192.168.0.20: icmp_seq=14 ttl=64 time=99.642 ms
64 bytes from 192.168.0.20: icmp_seq=15 ttl=64 time=77.294 ms
64 bytes from 192.168.0.20: icmp_seq=16 ttl=64 time=0.887 ms
64 bytes from 192.168.0.20: icmp_seq=17 ttl=64 time=1.978 ms
64 bytes from 192.168.0.20: icmp_seq=18 ttl=64 time=1.012 ms
64 bytes from 192.168.0.20: icmp_seq=19 ttl=64 time=4.542 ms

--- 192.168.0.20 ping statistics ---
20 packets transmitted, 20 packets received, 0.0% packet loss
round-trip min/avg/max/stddev = 0.887/18.933/99.642/30.483 ms

This server was a Raspberry Pi connected over gigabit ethernet to the modem/router.

Once again, notice the spikes.

It's not unusual for longer ping tests to have 5-10% packet loss, either.

What could be causing these symptoms? I'm almost sure it's the modem/router; I've tried connecting a downstream router broadcasting a different network, but the issue persists both for local and non-local hosts.

Jules

Posted 2016-10-20T07:04:45.713

Reputation: 628

Answers

3

Hmm, Seems like your wireless signals are getting affected. You can check for the following things:-

  • Whether there is any interference, check for any radio waves transmitting device. Be it a microwave, a bluetooth device or any other device transmitting on 5GHz band. Switch off any wireless device transmitting radio waves and then check.
  • Try changing the channels on 5GHz or even the band to 2.4GHz and see if that improves the latency.
  • Check whether you are getting strong signals from another wireless system, which may be transmitting on the same band(and channels) as yours and hence maybe affecting your SNR(Signal to Noise ratio).
  • You can also try and increase the transmitted power, See if that will improve the SNR and eventually the latency.
  • Updating the router version also might help.

Hope this helps!

Anirudh Malhotra

Posted 2016-10-20T07:04:45.713

Reputation: 664

Having done a scan, both my 5GHz and 2.4GHz bands were sharing channels with... about 10 other broadcast points. (Not really surprising, since this is an apartment). Changing the to less congested parts of the spectrum definitely had a big effect. The spikes are still there, but they're much less noticeable in SSH sessions. Out of curiosity, how could I go about boosting the signal strength? – Jules – 2016-10-20T14:18:14.843

I followed this link for you but didn't find anything on how to change the transmitted power I guess router automatically transmits at maximum power lever, You can see the transmitted power on DOCSIS WAN Screen though according to the document. What I also found was a way to automatically select channels(channels would switch when getting interfered, I guess). Hope this helps!

– Anirudh Malhotra – 2016-10-21T05:24:01.177