[Snap] Latency up a fair bit in the evenings; Congestion or something else

Forums › 2degrees (including Slingshot, Orcon, Flip, Stuff Fibre, MyRepublic, 2talk and Vocus) › [Snap] Latency up a fair bit in the evenings; Congestion or something else

1 | 2 | 3 | 4 | 5

68 posts

Master Geek

#714487 9-Nov-2012 08:20

here is a 8:20am 09/11 ping

Pinging 202.37.101.1 with 32 bytes of data:
Reply from 202.37.101.1: bytes=32 time=10ms TTL=61
Reply from 202.37.101.1: bytes=32 time=10ms TTL=61
Reply from 202.37.101.1: bytes=32 time=10ms TTL=61
Reply from 202.37.101.1: bytes=32 time=11ms TTL=61
Reply from 202.37.101.1: bytes=32 time=10ms TTL=61
Reply from 202.37.101.1: bytes=32 time=10ms TTL=61

will do a recheck this evening :)

vita

68 posts

Master Geek

#714900 9-Nov-2012 20:40

Seems much better now

20:40 09/11
Pinging 202.37.101.1 with 32 bytes of data:
Reply from 202.37.101.1: bytes=32 time=11ms TTL=61
Reply from 202.37.101.1: bytes=32 time=10ms TTL=61
Reply from 202.37.101.1: bytes=32 time=10ms TTL=61
Reply from 202.37.101.1: bytes=32 time=12ms TTL=61
Reply from 202.37.101.1: bytes=32 time=10ms TTL=61
Reply from 202.37.101.1: bytes=32 time=12ms TTL=61

Noticeably better in general browsing

mercutio

1392 posts

Uber Geek

#714902 9-Nov-2012 20:44

vita: Seems much better now

20:40 09/11
Pinging 202.37.101.1 with 32 bytes of data:
Reply from 202.37.101.1: bytes=32 time=11ms TTL=61
Reply from 202.37.101.1: bytes=32 time=10ms TTL=61
Reply from 202.37.101.1: bytes=32 time=10ms TTL=61
Reply from 202.37.101.1: bytes=32 time=12ms TTL=61
Reply from 202.37.101.1: bytes=32 time=10ms TTL=61
Reply from 202.37.101.1: bytes=32 time=12ms TTL=61

Noticeably better in general browsing

you're on the "high" ping non-interleaved ping rate for non-congestion, i've noticed some people seem to have like 6 msec pings on adsl. whereas other people are more ilke 9/10 ? .. i'm also on more like 9/10 ping.. does anyone know what causes this?

vita

68 posts

Master Geek

#715047 10-Nov-2012 09:56

Im not to worried about the ping atm - my total throughput speed is a bit on the low

9:53am 10/11

Pinging 202.37.101.1 with 32 bytes of data:
Reply from 202.37.101.1: bytes=32 time=11ms TTL=61
Reply from 202.37.101.1: bytes=32 time=69ms TTL=61
Reply from 202.37.101.1: bytes=32 time=90ms TTL=61
Reply from 202.37.101.1: bytes=32 time=34ms TTL=61
Reply from 202.37.101.1: bytes=32 time=12ms TTL=61
Reply from 202.37.101.1: bytes=32 time=22ms TTL=61
Reply from 202.37.101.1: bytes=32 time=83ms TTL=61

(dont know how to get the fancy speedtest.net image)
Date IP Address Download Upload Latency Server Distance
11/9/2012 8:53 PM GMT 111.69.202.66 5.81 Mb/s 0.68 Mb/s 33 ms Auckland < 50 m
11/9/2012 8:54 PM GMT 111.69.202.66 6.84 Mb/s 0.74 Mb/s 31 ms Christchurch ~ 450 mi

astrial

29 posts

Geek

#715888 12-Nov-2012 14:46

Hi, any further update on possible causes?

I was just browsing through the forums and stumbled across possibly a similar issue for Telecom:

http://www.geekzone.co.nz/forums.asp?forumid=39&topicid=111630

sidefx

3711 posts

Uber Geek

Trusted

#715890 12-Nov-2012 14:52

Yes, I'm watching that with interest too.

It appears to be somewhat improved from my observations, though whatever the issue is is still present. Whereas mine was up to high 40s in the evenings it now looks like it's down to ~high 30s. This appears to be backed up be my Truenet graphs which have dropped a bit, but are still showing a bit of a nasty peak from 8pm-11pm. I'll check again tonight manually.

"I was born not knowing and have had only a little time to change that here and there." | Octopus Energy | Sharesies
- Richard Feynman

Ragnor

8218 posts

Uber Geek

Trusted

#715901 12-Nov-2012 15:23

mercutio:

i've noticed some people seem to have like 6 msec pings on adsl. whereas other people are more ilke 9/10 ? .. i'm also on more like 9/10 ping.. does anyone know what causes this?

Number of hops between your cabinet/exchange and the regional aggregation point where Chorus hands over to your ISP's network would be the biggest factor.

Some ISP's have lots of handover points all round the country, some handover only in Auckland.

Snap used to only handover in Christchurch so North Island customers had to be back-hauled all the way back to Christchurch, they built an Auckland POP last year or year before.

Equinox IT

Cloud spending continues to surge globally, but most organisations haven’t made the changes necessary to maximise the value and cost-efficiency benefits of their cloud investments. Download the whitepaper From Overspend to Advantage now.

mercutio

1392 posts

Uber Geek

#715904 12-Nov-2012 15:28

Ragnor:
mercutio:

i've noticed some people seem to have like 6 msec pings on adsl. whereas other people are more ilke 9/10 ? .. i'm also on more like 9/10 ping.. does anyone know what causes this?

Number of hops between your cabinet/exchange and the regional aggregation point where your Chorus hands over to your ISP's network would be the biggest factor.

Some ISP's have lots of handover points all round the country, some handover only in Auckland.

Snap used to only handover in Christchurch so North Island customers had to be back-hauled all the way back to Christchurch, they built an Auckland POP last year or year before.

yeah, but i mean being a suburb over shouldn't make 3 or 4 msec difference should it?

from what i can tell there's slightly higher latency on cabinets then exchanges. then is it just that some exchanges are better connected than others? i know snap only have handover in auckland and christchurch, but that shouldn't make much difference in either of those regions?

Ragnor

8218 posts

Uber Geek

Trusted

#715907 12-Nov-2012 15:36

mercutio:

yeah, but i mean being a suburb over shouldn't make 3 or 4 msec difference should it?

It depends.. Are they on the same ISP? Same base connection service EUBA vs BUBA?

Neighbouring suburbs could easily be in different exchange area's, with different activity/congestion levels etc.

A few ms difference is nothing at the end of the day and residential DSL is best effort.

mercutio

1392 posts

Uber Geek

#715909 12-Nov-2012 15:46

Ragnor:
mercutio:

yeah, but i mean being a suburb over shouldn't make 3 or 4 msec difference should it?

It depends.. Are they on the same ISP? Same base connection service EUBA vs BUBA?

Neighbouring suburbs could easily be in different exchange area's, with different activity/congestion levels etc.

A few ms difference is nothing at the end of the day and residential DSL is best effort.

well, i'm using slingshot, snap, and orcon for comparison's sake, and the lowest latency has been snap euba/exchange followed by slingshot llu / orcon euba/exchange, and the highest latency has been snap euba/cabinet followed by orcon euba/cabinet.

and comparing two lines that have switched from cabinet to exchange, both went up in latency. i realise it's best effort, and only makes a slight difference, and that the difference been euba low interleaving and no interleaving is more noticable. i'm mostly just curious how latency ends up being higher.

in earlier tests i seem to recall interleaving off was about 10% faster for web browsing average page load times than interleaving low. and about 20% faster than interleaving on. so it's probably only a few % difference in the real world.

that said that's in a higher parallel activity such as web browsing, and things like NFS (network file system) tend to be much more serial and latency-bound.

RalphFromSnap

776 posts

Ultimate Geek

Trusted

Snap Internet

#716057 12-Nov-2012 20:17

Hi All,

We currently have this under investigation by Chorus, we expect that the impact may be wider than just Snap customers.

In regard to the EUBA/BUBA question, all Snap customers on ADSL2/ISAM's are automatically transferred to EUBA by our provisioning system as soon as it detects that they are on capable hardware

Also we have multiple handovers over the country for ADSL2/VDSL (and HSNS Business services) in the South Island in all towns where this is available, and in the North Island in many of the main centres and multiple in Auckland. However on ADSL1/BUBA you will either be handed over in Auckland or Christchurch, generally routed to the North.

Cheers.

Snap

0800 BROADBAND (276 232)
www.snap.net.nz

@SnapInternet on Twitter
Snap Internet on Facebook

Our Social Media Team:
^RO Ricky - Technical Lead
^AT Ashleigh - Retail Marketing Coordinator

sidefx

3711 posts

Uber Geek

Trusted

#719202 18-Nov-2012 21:43

My latency is looking very good tonight:

18/11/2012 20:27:41.67

Pinging 202.37.101.1 with 32 bytes of data:
Reply from 202.37.101.1: bytes=32 time=6ms TTL=61
Reply from 202.37.101.1: bytes=32 time=5ms TTL=61
Reply from 202.37.101.1: bytes=32 time=5ms TTL=61
Reply from 202.37.101.1: bytes=32 time=5ms TTL=61
Reply from 202.37.101.1: bytes=32 time=5ms TTL=61
Reply from 202.37.101.1: bytes=32 time=5ms TTL=61
Reply from 202.37.101.1: bytes=32 time=5ms TTL=61
Reply from 202.37.101.1: bytes=32 time=5ms TTL=61
Reply from 202.37.101.1: bytes=32 time=5ms TTL=61
Reply from 202.37.101.1: bytes=32 time=5ms TTL=61

Ping statistics for 202.37.101.1:
Packets: Sent = 10, Received = 10, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 5ms, Maximum = 6ms, Average = 5ms

...

18/11/2012 21:00:04.31

Pinging 202.37.101.1 with 32 bytes of data:
Reply from 202.37.101.1: bytes=32 time=5ms TTL=61
Reply from 202.37.101.1: bytes=32 time=6ms TTL=61
Reply from 202.37.101.1: bytes=32 time=5ms TTL=61
Reply from 202.37.101.1: bytes=32 time=5ms TTL=61
Reply from 202.37.101.1: bytes=32 time=5ms TTL=61
Reply from 202.37.101.1: bytes=32 time=5ms TTL=61
Reply from 202.37.101.1: bytes=32 time=5ms TTL=61
Reply from 202.37.101.1: bytes=32 time=5ms TTL=61
Reply from 202.37.101.1: bytes=32 time=5ms TTL=61
Reply from 202.37.101.1: bytes=32 time=5ms TTL=61

Ping statistics for 202.37.101.1:
Packets: Sent = 10, Received = 10, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 5ms, Maximum = 6ms, Average = 5ms

...

18/11/2012 21:30:39.31

Pinging 202.37.101.1 with 32 bytes of data:
Reply from 202.37.101.1: bytes=32 time=5ms TTL=61
Reply from 202.37.101.1: bytes=32 time=5ms TTL=61
Reply from 202.37.101.1: bytes=32 time=5ms TTL=61
Reply from 202.37.101.1: bytes=32 time=5ms TTL=61
Reply from 202.37.101.1: bytes=32 time=5ms TTL=61
Reply from 202.37.101.1: bytes=32 time=5ms TTL=61
Reply from 202.37.101.1: bytes=32 time=6ms TTL=61
Reply from 202.37.101.1: bytes=32 time=5ms TTL=61
Reply from 202.37.101.1: bytes=32 time=5ms TTL=61
Reply from 202.37.101.1: bytes=32 time=5ms TTL=61

Ping statistics for 202.37.101.1:
Packets: Sent = 10, Received = 10, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 5ms, Maximum = 6ms, Average = 5ms

...

18/11/2012 21:40:19.31

Pinging 202.37.101.1 with 32 bytes of data:
Reply from 202.37.101.1: bytes=32 time=5ms TTL=61
Reply from 202.37.101.1: bytes=32 time=5ms TTL=61
Reply from 202.37.101.1: bytes=32 time=5ms TTL=61
Reply from 202.37.101.1: bytes=32 time=6ms TTL=61
Reply from 202.37.101.1: bytes=32 time=5ms TTL=61
Reply from 202.37.101.1: bytes=32 time=5ms TTL=61
Reply from 202.37.101.1: bytes=32 time=5ms TTL=61
Reply from 202.37.101.1: bytes=32 time=5ms TTL=61
Reply from 202.37.101.1: bytes=32 time=5ms TTL=61
Reply from 202.37.101.1: bytes=32 time=5ms TTL=61

Ping statistics for 202.37.101.1:
Packets: Sent = 10, Received = 10, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 5ms, Maximum = 6ms, Average = 5ms

Trademe looking good too:

PS C:\> ping /t 202.162.73.2

Pinging 202.162.73.2 with 32 bytes of data:
Reply from 202.162.73.2: bytes=32 time=7ms TTL=250
Reply from 202.162.73.2: bytes=32 time=6ms TTL=250
Reply from 202.162.73.2: bytes=32 time=6ms TTL=250
Reply from 202.162.73.2: bytes=32 time=6ms TTL=250
Reply from 202.162.73.2: bytes=32 time=6ms TTL=250
Reply from 202.162.73.2: bytes=32 time=6ms TTL=250

Ping statistics for 202.162.73.2:
Packets: Sent = 6, Received = 6, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 6ms, Maximum = 7ms, Average = 6ms
Control-C
PS C:\>

"I was born not knowing and have had only a little time to change that here and there." | Octopus Energy | Sharesies
- Richard Feynman

bender

220 posts

Master Geek

#719248 19-Nov-2012 06:02

Yep Chorus have confirmed the issue has been resolved

sidefx

3711 posts

Uber Geek

Trusted

#721025 21-Nov-2012 22:42

Yep, after a few days to average out my truenet graphs (and manual ping tests to various sites) are back to normal, looking good for all times of day!

Just out of curiosity can anyone offer any further info on what the cause was?

"I was born not knowing and have had only a little time to change that here and there." | Octopus Energy | Sharesies
- Richard Feynman

bender

220 posts

Master Geek

#721027 21-Nov-2012 22:46

There was congestion on one of the Ethernet links that goes from Mt Albert to Mayoral Drive. Anyone on an exchange that aggregates at Mt Albert (eg Three Kings) was affected if they were going over the affected link.

1 | 2 | 3 | 4 | 5