dslreports logo
site
 
    All Forums Hot Topics Gallery
spc

spacer




how-to block ads


Search Topic:
uniqs
920
share rss forum feed

dtherbert

join:2009-04-24
Evanston, IL
reply to dtherbert

Re: [Connectivity] Latency Issue

Click for full size
I ran PingPlotter for 24 hours, anybody see anything wrong with this picture?

The really bad looking part is the only time a load was present on the connection. It wasn't even that much of a load. We have a 20/4 connection, and we were only watching one Netflix HD stream and browsing the web, so maybe 2-10 Mb/s at any given time. This resulted in huge latency spikes.

As an aside, how does my connection look at all other times of the day?

If I ping the first hop after the router, that's where all the spikes come from it seems, so I think that's where the problems are. This is really making online gaming unenjoyable, so any information as to what the problem is and exactly what I can do about it would be greatly appreciated. Thank you!

dtherbert

join:2009-04-24
Evanston, IL
reply to devnuller

Thanks devnuller, I'll check that out.

Also, I've been running smokeping for awhile now, see link:

»/r3/smokeping.···99f40735

Anybody care to interpret the results?

I'm definitely seeing some packet loss, on the order of a few percentage points. Also, the period on the graphs corresponding to 8:30PM-midnight, I was streaming Netflix for a good portion of that time, and it seems to have dropped a few more packets, and the pings were definitely less consistent (I probably should have prioritized ICMP packets on my router).


devnuller

join:2006-06-10
Cambridge, MA
reply to dtherbert

said by dtherbert:

Is there a recommendation for a Mac-compatible alternative to PingPlotter?

mtr - »www.bitwizard.nl/mtr/


NormanS
I gave her time to steal my mind away
Premium,MVM
join:2001-02-14
San Jose, CA
kudos:11
Reviews:
·SONIC.NET
reply to dtherbert

said by dtherbert:

Is there a recommendation for a Mac-compatible alternative to PingPlotter?

Maybe Smokeping, though I don't know how to interpret the results.
--
Norman
~Oh Lord, why have you come
~To Konnyu, with the Lion and the Drum

dtherbert

join:2009-04-24
Evanston, IL
reply to NormanS

Is there a recommendation for a Mac-compatible alternative to PingPlotter?



NormanS
I gave her time to steal my mind away
Premium,MVM
join:2001-02-14
San Jose, CA
kudos:11
Reviews:
·SONIC.NET
reply to dtherbert

Many routers are programmed to give low priority to diagnostic packets directed at them. Their primary job is forwarding TCP/IP traffic up the path. Response to diagnostics, unlike routing through, requires processor time; which may cause throughput delays if the router is 100% responsive to ICMP, and the like.
--
Norman
~Oh Lord, why have you come
~To Konnyu, with the Lion and the Drum


dtherbert

join:2009-04-24
Evanston, IL
reply to NormanS

I will do these things you suggest.

To answer your question about pinging an intermediary router, I thought I would try pinging the addresses listed in the trace, starting with my router, then the modem (not listed, at 192.168.100.1), then beyond, to see if I got spikes anywhere. It happened that the first address listed after my router showed these ping spikes, so I didn't try any further ones since I thought that was where the problem was. Isn't the fact that I'm getting a response mean that it is responding to ICMP, or is my reasoning flawed (timeouts/spikes due to low-priority or something like that)?

Anyway, thank you for the response, I'll post the new information after I get home later.



NormanS
I gave her time to steal my mind away
Premium,MVM
join:2001-02-14
San Jose, CA
kudos:11
Reviews:
·SONIC.NET
reply to dtherbert

Why are you pinging an intermediary router, which is likely going to be unresponsive to ICMP?

How about pinging Google for 50 iterations; and just post the summary, because that has all the information that you need to show:

Ping statistics for 74.125.224.146:
    Packets: Sent = 50, Received = 50, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
    Minimum = 24ms, Maximum = 26ms, Average = 25ms
 
I don't think your signals show a problem, but I will defer to any DOCSIS experts, if they disagree.

Your trace doesn't look bad, but it is a single point in time. A PingPlot might show something.

PingPlot to Google over time.


--
Norman
~Oh Lord, why have you come
~To Konnyu, with the Lion and the Drum

dtherbert

join:2009-04-24
Evanston, IL

Hi there,

I'm having really large latency spikes, on the order of 1-3 seconds. If I ping www.google.com continuously from the terminal on my Mac, I'll have pretty steady responses for maybe 15-50 reps, followed by a timeout or two, then a couple of pings that return 1-3 seconds later. Repeat.

I did a traceroute to google and got:

traceroute to www.google.com (74.125.225.176), 64 hops max, 52 byte packets
1 192.168.1.1 (192.168.1.1) 3.816 ms 1.837 ms 1.061 ms
2 c-24-12-184-1.hsd1.il.comcast.net (24.12.184.1) 26.248 ms 19.518 ms 15.572 ms
3 te-6-7-ur02.mortongrove.il.chicago.comcast.net (68.85.208.221) 13.346 ms 10.419 ms 13.811 ms
4 te-1-4-0-0-ar01.area4.il.chicago.comcast.net (68.87.230.65) 20.090 ms 17.520 ms 15.870 ms
5 he-3-10-0-0-cr01.350ecermak.il.ibone.comcast.net (68.86.93.181) 16.482 ms 17.865 ms 22.736 ms
6 * * pos-1-5-0-0-pe01.350ecermak.il.ibone.comcast.net (68.86.87.126) 20.033 ms
7 66.208.228.202 (66.208.228.202) 65.570 ms * *
8 209.85.254.128 (209.85.254.128) 23.450 ms
209.85.254.120 (209.85.254.120) 22.895 ms
209.85.254.128 (209.85.254.128) 18.657 ms
9 72.14.237.110 (72.14.237.110) 122.742 ms 103.439 ms 103.051 ms
10 209.85.241.22 (209.85.241.22) 25.606 ms 28.368 ms
72.14.232.141 (72.14.232.141) 25.425 ms
11 72.14.239.51 (72.14.239.51) 39.147 ms 36.704 ms 55.826 ms
12 216.239.46.145 (216.239.46.145) 34.121 ms 37.407 ms 39.493 ms
13 209.85.251.9 (209.85.251.9) 37.259 ms 35.837 ms 43.547 ms
14 den03s05-in-f16.1e100.net (74.125.225.176) 36.221 ms 38.430 ms 37.253 ms

Now, if I ping the router or the modem (192.168.100.1, it's an SB6121), I'll get consistant pings until the cows come home. It's when I ping that hop to 24.12.184.1 that I get the spikes. See below:

PING 24.12.184.1 (24.12.184.1): 56 data bytes
64 bytes from 24.12.184.1: icmp_seq=0 ttl=254 time=12.387 ms
64 bytes from 24.12.184.1: icmp_seq=1 ttl=254 time=13.955 ms
Request timeout for icmp_seq 2
Request timeout for icmp_seq 3
64 bytes from 24.12.184.1: icmp_seq=2 ttl=254 time=2755.662 ms
64 bytes from 24.12.184.1: icmp_seq=3 ttl=254 time=1755.403 ms
64 bytes from 24.12.184.1: icmp_seq=4 ttl=254 time=755.224 ms
64 bytes from 24.12.184.1: icmp_seq=5 ttl=254 time=17.830 ms
64 bytes from 24.12.184.1: icmp_seq=6 ttl=254 time=14.863 ms
64 bytes from 24.12.184.1: icmp_seq=7 ttl=254 time=17.343 ms
64 bytes from 24.12.184.1: icmp_seq=8 ttl=254 time=19.307 ms
64 bytes from 24.12.184.1: icmp_seq=9 ttl=254 time=10.825 ms
64 bytes from 24.12.184.1: icmp_seq=10 ttl=254 time=1162.883 ms
64 bytes from 24.12.184.1: icmp_seq=11 ttl=254 time=163.342 ms
64 bytes from 24.12.184.1: icmp_seq=12 ttl=254 time=13.765 ms
64 bytes from 24.12.184.1: icmp_seq=13 ttl=254 time=15.521 ms
64 bytes from 24.12.184.1: icmp_seq=14 ttl=254 time=11.994 ms
64 bytes from 24.12.184.1: icmp_seq=15 ttl=254 time=11.034 ms
64 bytes from 24.12.184.1: icmp_seq=16 ttl=254 time=14.339 ms
64 bytes from 24.12.184.1: icmp_seq=17 ttl=254 time=11.608 ms
64 bytes from 24.12.184.1: icmp_seq=18 ttl=254 time=21.548 ms
64 bytes from 24.12.184.1: icmp_seq=19 ttl=254 time=14.356 ms
64 bytes from 24.12.184.1: icmp_seq=20 ttl=254 time=14.228 ms
64 bytes from 24.12.184.1: icmp_seq=21 ttl=254 time=15.874 ms
64 bytes from 24.12.184.1: icmp_seq=22 ttl=254 time=13.354 ms
64 bytes from 24.12.184.1: icmp_seq=23 ttl=254 time=13.983 ms
64 bytes from 24.12.184.1: icmp_seq=24 ttl=254 time=13.393 ms
64 bytes from 24.12.184.1: icmp_seq=25 ttl=254 time=13.357 ms
64 bytes from 24.12.184.1: icmp_seq=26 ttl=254 time=14.691 ms
64 bytes from 24.12.184.1: icmp_seq=27 ttl=254 time=13.011 ms
64 bytes from 24.12.184.1: icmp_seq=28 ttl=254 time=16.891 ms
64 bytes from 24.12.184.1: icmp_seq=29 ttl=254 time=16.980 ms
64 bytes from 24.12.184.1: icmp_seq=30 ttl=254 time=1779.668 ms
64 bytes from 24.12.184.1: icmp_seq=31 ttl=254 time=779.454 ms
64 bytes from 24.12.184.1: icmp_seq=32 ttl=254 time=12.508 ms
64 bytes from 24.12.184.1: icmp_seq=33 ttl=254 time=9.860 ms
64 bytes from 24.12.184.1: icmp_seq=34 ttl=254 time=15.418 ms
64 bytes from 24.12.184.1: icmp_seq=35 ttl=254 time=13.819 ms

So, any ideas about how to handle this?

Also, signal levels for those interested:

Channel ID 2 3 6 8
Frequency 657000000 Hz 663000000 Hz 681000000 Hz 693000000 Hz
Signal to Noise Ratio 37 dB 37 dB 37 dB 37 dB
Downstream Modulation QAM256 QAM256 QAM256 QAM256
Power Level
The Downstream Power Level reading is a snapshot taken at the time this page was requested. Please Reload/Refresh this Page for a new reading
0 dBmV 0 dBmV 0 dBmV 0 dBmV
Upstream Bonding Channel Value
Channel ID 8 7 9
Frequency 29500000 Hz 36400000 Hz 24200000 Hz
Ranging Service ID 4741 4741 4741
Symbol Rate 5.120 Msym/sec 5.120 Msym/sec 2.560 Msym/sec
Power Level 48 dBmV 48 dBmV 47 dBmV
Upstream Modulation [3] QPSK
[3] 64QAM
[3] QPSK
[3] 64QAM
[3] QPSK
[2] 16QAM

Ranging Status Success Success Success
Signal Stats (Codewords) Bonding Channel Value
Channel ID 2 3 6 8
Total Unerrored Codewords 237152973 236462702 236462618 236462972
Total Correctable Codewords 0 2 0 0
Total Uncorrectable Codewords 515 588 566 508