Shaw → High latency/ping to Shaw router?

Hey everyone! I'm new here :)
Firstly, I'm not sure if this is the best forum for this post, please let me know if it isn't!

I don't know if this is recent (probably not), but I've been noticing relatively high latency (20ms) between my endpoint and the first Shaw router I can see in the traceroute.
My internet speeds are okay (70Mbps/3Mbps on Internet 60).

I'm wondering if this is standard latency? I know that from UBC (on the UBC network), ping to a downtown router is only a couple of mills, and ping to SFO from UBC is around 20ms.

I live on the UBC campus.

Here's a traceroute from my home network to OpenDNS Vancouver:

traceroute to resolver1.opendns.com (208.67.222.222), 64 hops max, 52 byte packets
 1  192.168.1.1 (192.168.1.1)  4.312 ms  7.829 ms  10.082 ms
 2  * * *
 3  rd3st-tge0-12-0-6-1.vc.shawcable.net (64.59.147.245)  32.389 ms  19.667 ms  32.647 ms
 4  vanix.opendns.com (206.41.104.40)  13.617 ms  14.408 ms  14.617 ms
 5  resolver1.opendns.com (208.67.222.222)  23.864 ms  19.308 ms  30.002 ms
 

And one from UBC to OpenDNS:

traceroute to resolver1.opendns.com (208.67.222.222), 64 hops max, 52 byte packets
 1  10.20.216.29 (10.20.216.29)  2.039 ms  1.398 ms  1.365 ms
 2  142.103.204.138 (142.103.204.138)  1.574 ms  1.853 ms  1.779 ms
 3  anguborder-a0.net.ubc.ca (137.82.123.137)  1.703 ms  1.520 ms  1.499 ms
 4  345-ix-cr1-ubcab.vncv1.bc.net (134.87.0.58)  2.069 ms  1.919 ms  3.206 ms
 5  vanix.opendns.com (206.41.104.40)  1.925 ms  2.142 ms  1.863 ms
 6  resolver1.opendns.com (208.67.222.222)  1.575 ms  1.672 ms  2.393 ms
 

Any suggestions, and anyone else having this issue (if it is an issue to begin with)?

Sounds like same issue I'm seeing with the new DPC3848V modem...

all 24 channel modems shaw is using sucks

since your on 60 you should be able to ask for DPC3825 model and problems should be resolved

tcp is not affected by it, so speedtest.net shows same low ping. icmp gets hit hard by this

checkout this post and mainly pay attention to my smokeping charts

»CGNM-2250 Problems (CGNM-2250 VS DPC3825)

however should this affect regular usage, not at all, gaming or latency dependent applications like voip will be negatively impacted by this, but 20ms is not very noticeable either way ( in sound the average person can not tell the difference in delays of 10ms or less) however drummers and some extraordinary people are able to get down finer then 1ms

Thanks for the info! I had seen on the forums some of the issues with the newer modems, but I already have a DPC3825 in bridged mode.

The high latency doesn't affect regular web activities as you mentioned, but things like gaming and remote work (remote desktop, SSH) are affected quite badly.

if your using DPC3825

first call shaw and nicely ask them what the usage levels are on the channels your modem is using and if they are below 65% i would begin to try to get your powerlevels on your modem better.

a higher ustream power output can cause the modem to run very hot and performance will drop

ideally you want all rx channels between + or - 6

and all tx channels 45 or lower

and if your SNR on your DS channels is lower then 35, your modem will also be working harder.

shaw will say blah blah blah + or - 15 (or maybe they say 12,but i think 12 is orange for them)

blah blah blah you dont need high snr

hey shaw, fuck you

IMO shaw techs need to get powerlevels to a better point then just "usable"

Hmm. I have a DPC3825 modem (in bridged mode).

well then if your having ping issues i would suspect node congestion but you said your getting full speed so i would next suggest TX/RX/SNR issues with your cable modem

Thanks for your help, Fryst! My replies were held up by the mod system so they didn't get out immediately, nor in the right order.

I'll definitely investigate further into what you said.

Thanks again!

Oh, one more thing I noticed that might be a factor -- I recently signed up for a phone bundle, so I now have a Shaw-installed splitter (to split for internet modem/phone modem).

It's identical to this one:

thats exactly the type of thing which would cause ur modem levels to get worse and make your ping less stable, however it should only make minor differences, unless your modem was going out of spec like past + or - 12 (or really low SNR)

if you wanna further diagnose

constant ping A) google

B) your modem 192.168.100.1 (try it it will work)

and C)prerferably default gateway for your routers ip, assuming you can find it, if not use 64.59.147.245

but try to find ur default gateway, and if you do i would also ping the above address

make sure no load is on ur network and watch

google/default gateway should only fluctuate 1-3ms (maybe 4 or 5 occasionally)

any more then that is no good IMO

That latency you see isn't real. How do I know? The next hop doesn't have the same or higher latency. All you are seeing is a Cisco big-iron router with a busy CPU that is rate limiting ICMP Time Exceeded message creation.

See slide 32/33 here: »www.nanog.org/meetings/n ··· _N45.pdf

Cisco routers respond to packet TTL expiry (how a traceroute is done) in software on the CPU, but your actual traffic is forwarded by dedicated ASICs on the interfaces, bypassing the CPU completely.

Traceroutes are becoming less and less interesting, a lot of fake latency results out there. Even modems now de-prioritize ICMP. If the latency doesn't continually progress up, it's fake latency.

You can't go from 30ms round-trip time on hop 3, to 14 ms on hop 4. How did that next hop go back in time 16 ms? The answer is that the packet expiry sat in the queue for the CPU for close to 15 ms before the ICMP time exceeded message was created and queued on the return interface.

Thanks for the info! Those set of slides are a good read.

Even considering that, there is still a considerable difference in latency (at least 10ms) between the two ISP's. This might be caused by my setup or by some queueing delay enroute with Shaw.

Going back to latency, how do you usually measure this accurately?

your point is valid

however his default gateway IP should respond without a delay, which is why i mentioned to find his default gateway

he is still getting at best 19-20ms to 208.67.222.222

his first hop time should be very similar to mine however im on telus and they might just have better routing to the server, 19-20ms does seem fine you are correct ravenchild

tracert 208.67.222.222

Tracing route to resolver1.opendns.com [208.67.222.222]
over a maximum of 30 hops:

1 1 ms 1 ms 1 ms EdgeRouterX [192.168.1.1]
2 7 ms 6 ms 6 ms 10.29.214.1
3 9 ms 9 ms 9 ms STTLWAWBCI01.bb.telus.com [75.154.217.108]
4 10 ms 9 ms 9 ms six.opendns.com [206.81.80.53]
5 10 ms 9 ms 9 ms resolver1.opendns.com [208.67.222.222]

Trace complete.

i hope your initial pings to the opendns servers was on LAN no WIFI

EDIT: i also just wanted to say i dont believe thats opendns vancouver its going through seattle it has seattle CLLI code and shows SIX

Here are some more tests if it's helpful (a variety of ping and traceroute):

Traceroute to modem:

traceroute to 192.168.100.1 (192.168.100.1), 64 hops max, 52 byte packets
 1  192.168.1.1 (192.168.1.1)  4.200 ms  4.184 ms  9.697 ms
 2  192.168.100.1 (192.168.100.1)  20.311 ms  9.593 ms  5.191 ms
 

Ping to modem:

PING 192.168.100.1 (192.168.100.1): 56 data bytes
64 bytes from 192.168.100.1: icmp_seq=0 ttl=63 time=15.136 ms
64 bytes from 192.168.100.1: icmp_seq=1 ttl=63 time=20.054 ms
64 bytes from 192.168.100.1: icmp_seq=2 ttl=63 time=4.921 ms
64 bytes from 192.168.100.1: icmp_seq=3 ttl=63 time=4.568 ms
64 bytes from 192.168.100.1: icmp_seq=4 ttl=63 time=5.978 ms
64 bytes from 192.168.100.1: icmp_seq=5 ttl=63 time=12.677 ms
64 bytes from 192.168.100.1: icmp_seq=6 ttl=63 time=9.129 ms
64 bytes from 192.168.100.1: icmp_seq=7 ttl=63 time=10.344 ms
64 bytes from 192.168.100.1: icmp_seq=8 ttl=63 time=10.220 ms
64 bytes from 192.168.100.1: icmp_seq=9 ttl=63 time=3.742 ms
 
--- 192.168.100.1 ping statistics ---
10 packets transmitted, 10 packets received, 0.0% packet loss
round-trip min/avg/max/stddev = 3.742/9.677/20.054/4.951 ms
 

And ping to default gateway (50.67.20.1). At least I think this is the right gateway. It's the one that shows up in my wireless router config (assigned via DHCP):

PING 50.67.20.1 (50.67.20.1): 56 data bytes
64 bytes from 50.67.20.1: icmp_seq=0 ttl=254 time=15.879 ms
64 bytes from 50.67.20.1: icmp_seq=1 ttl=254 time=20.341 ms
64 bytes from 50.67.20.1: icmp_seq=2 ttl=254 time=25.594 ms
64 bytes from 50.67.20.1: icmp_seq=3 ttl=254 time=20.437 ms
64 bytes from 50.67.20.1: icmp_seq=4 ttl=254 time=15.289 ms
64 bytes from 50.67.20.1: icmp_seq=5 ttl=254 time=19.382 ms
64 bytes from 50.67.20.1: icmp_seq=6 ttl=254 time=17.477 ms
64 bytes from 50.67.20.1: icmp_seq=7 ttl=254 time=23.431 ms
64 bytes from 50.67.20.1: icmp_seq=8 ttl=254 time=15.470 ms
64 bytes from 50.67.20.1: icmp_seq=9 ttl=254 time=14.441 ms
 
--- 50.67.20.1 ping statistics ---
10 packets transmitted, 10 packets received, 0.0% packet loss
round-trip min/avg/max/stddev = 14.441/18.774/25.594/3.545 ms
 

I'll talk to Shaw and play around with the splitter when I have a bit more time (maybe tomorrow or over the weekend).It seems like, in your traceroute it's going to Seattle, but in my case, it's ending up in Vancouver (vanix.opendns.com). OpenDNS uses an anycast system so it's definitely possible, but I'm not sure why Telus is sending you to their Seattle node. I'm super jealous of that ping though!

Update: I just tried without splitter, the traceroute times did not noticeably change.

??? how can 208.67.222.222 be in 2 places ? that doesnt make any sense to me??

and dude, are u on WIFI? if ur using wifi you cant count on anything to be steady

Looking at your first hop, you are on WiFi or whatever is at 192.168.1.1 has issues.

Until that is showing 1ms or less, no point in really looking anywhere else.

Anycast (the same IP address/prefix is announced by multiple routers in multiple locations, so the end user gets routed to the nearest one): »en.wikipedia.org/wiki/AnycastNo, this is on ethernet, I swear!

This is my ping from my computer (newly-built desktop with Intel gigabit ethernet port), over ethernet, to my wireless router (Linksys E2500):

PING 192.168.1.1 (192.168.1.1): 56 data bytes
64 bytes from 192.168.1.1: icmp_seq=0 ttl=64 time=10.361 ms
64 bytes from 192.168.1.1: icmp_seq=1 ttl=64 time=0.958 ms
64 bytes from 192.168.1.1: icmp_seq=2 ttl=64 time=6.191 ms
64 bytes from 192.168.1.1: icmp_seq=3 ttl=64 time=1.455 ms
64 bytes from 192.168.1.1: icmp_seq=4 ttl=64 time=4.895 ms
64 bytes from 192.168.1.1: icmp_seq=5 ttl=64 time=3.604 ms
64 bytes from 192.168.1.1: icmp_seq=6 ttl=64 time=2.425 ms
64 bytes from 192.168.1.1: icmp_seq=7 ttl=64 time=7.598 ms
64 bytes from 192.168.1.1: icmp_seq=8 ttl=64 time=1.163 ms
64 bytes from 192.168.1.1: icmp_seq=9 ttl=64 time=1.133 ms
 
--- 192.168.1.1 ping statistics ---
10 packets transmitted, 10 packets received, 0.0% packet loss
round-trip min/avg/max/stddev = 0.958/3.978/10.361/3.060 ms
 

You are of course welcome to help diagnose this as well ;)

that looks like a problem with your router.....

that is so wierd i cant ping 206.41.104.40 ?? its like i only get access to it if i get routed to the vancovuer opendns, now im thinking this is a little issue on telus end

C:\Users\Connor>ping 206.41.104.40

Pinging 206.41.104.40 with 32 bytes of data:
Request timed out.
Request timed out.

Ping statistics for 206.41.104.40:
Packets: Sent = 2, Received = 0, Lost = 2 (100% loss),
Control-C
^C

tracert 206.41.104.40

Tracing route to vanix.opendns.com [206.41.104.40]
over a maximum of 30 hops:

1 5 ms ping vanix.opendns.com
Ping request could not find host vanix.opendns.com. Please check

Anything I can do to find the cause of the issue? (check ethernet cable, router configuration?)

Is the jitter to my router bad enough to contribute significantly to my high latencies? The ping to router only differs by 1-5ms.

what router are you using?

Don't worry, I can't directly ping it either. Probably some firewall thing blocking pings? vanix.opendns.com is the PTR (reverse lookup) entry for that address, but it doesn't seem resolve the other way around (no A record from hostname to IP)

I wouldn't worry too much about it, perhaps it's just that it really is a shorter path from Telus to the SIX node.

It's a Linksys E2500. (»www.linksys.com/us/suppo ··· 3KZuNAAW)

i wouldnt count on linksys firmware for the life of me so i would try tomato out, here is a link to where u can get a E2500 build

»tomato.groov.pl/download ··· -series/

i would find out if you have a v1,v2, or v3

v1 and v2 use the same FW v3 uses a different one (v1,v2 have 8MB flash v3 has 16MB)

»tomato.groov.pl/download ··· -series/

you can upgrade directy from the linksys firmware just go to the upgrade firmware and link it to the tomato .bin file inside the zip you download

I did think of using a third-party firmware, but was skeptical about whether it would work better than stock firmware (since the stock is "optimized" specifically for the hardware?) Of course this is all mere speculation on my part.

Would you recommend Tomato over DD-WRT?

yes, i have tried both, to me tomato makes more sense

i highly recommend trying tomato

Thanks! I will try it out

good luck

So, I installed Tomato, and ping times have gone down quite a bit!

Now I'm seeing approximately 11-12ms latency to downtown:

traceroute to resolver1.opendns.com (208.67.222.222), 64 hops max, 52 byte packets
 1  192.168.1.1 (192.168.1.1)  0.720 ms  0.540 ms  0.561 ms
 2  * * *
 3  rd3st-tge0-12-0-6-1.vc.shawcable.net (64.59.147.245)  12.389 ms  11.955 ms  10.899 ms
 4  vanix.opendns.com (206.41.104.40)  12.741 ms  10.467 ms  11.228 ms
 5  resolver1.opendns.com (208.67.222.222)  11.348 ms  12.145 ms  16.150 ms
 

Latency to the router is low now too, sub-millisecond.

Thanks for all of your help! Do you still think I should contact Shaw to make sure my modem settings are okay?