Geekzone: technology news, blogs, forums
Guest
Welcome Guest.
You haven't logged in yet. If you don't have an account you can register now.

View this topic in a long page with up to 500 replies per page Create new topic
1 | 2 | 3 | 4 | 5
1078 posts

Uber Geek
+1 received by user: 45


  Reply # 702253 16-Oct-2012 21:57 Send private message

Stasis007: Maybe related, maybe not. I've had spells of horrible latency in World of Warcraft the last week or two. Usually happens between 8-11pm. Worst I've seen in over 3 years, etc. Actually DC's me often, and local pings are fine but 2k+ pings to eastern USA.

Mate on Telstra was experiencing exactly the same which seems significant.


FWIW, it seems that Diablo 3 has had some latency spikes around 7 pm recently.   (haven't played it much) and around the same time cogent interim hops seem to show packet loss.

The 203.167. clear address in the middle often shows high ping times but this doesn't seem to translate to bad performance, but the cogent and at&t ones do.

Really if you're playing wow you should use west coast servers, but I'd try using pingplotter to the destination server which you can find in "resource monitor" under network.

If jitter (time difference between minimum and average ping) is more than 10 msec near the destination, or there is packet loss showing on pingplotter it could be noticable.  The thing about playing games is there's a constant stream of packets - so if you get 1% packet loss, and the game sends 10,000 packets then you're actually missing 100 packets, which could lead to 100 spikes.  Really blizzard should fix their game to deal better with packet loss, but if you look around on the internet issues are reasonably common over whole isp's at various times, as blizzard use a provider who charge for access to their network.. (kind of liek if you hosted a game on telecom, and lots of isp's got peak congestion to telecom, they'd have to buy more transit to telecom to raise performance, adn then who should you blame, the company who decides to host on telecom, telecom for expecting others to pay to receive data from them, or ISP's not wishing to pay for direct transit. 

Right now as it stands most ISP's buy national traffic to telstraclear or globalgateway, and not both seperately.  So there'll be a longer router to providers using globalgateway for domestic traffic - but it's not generally oversubscribed in NZ.

Anyway, as it stands I'm not aware of any provider in NZ having direct connectivity to at&t, and you're relying on cogent for connectivity to at&t atm which have a reputation for oversubscribing transit links.

Although currently there aren't strong national peering issues nationally there a few between NZ<->Australia, where providers are slightly more keen to bypass the tradtional telco's.


29 posts

Geek


  Reply # 702643 17-Oct-2012 19:25 Send private message

Here's mine, pinged just now.  I was actually going to log in and say how awesome my ping has been to US West DOTA2 servers (reporting ~160ms in game generally across night peak), but I see here that the jitter to the Snap DNS server is pretty high!

Pinging 202.37.101.1 with 32 bytes of data:
Reply from 202.37.101.1: bytes=32 time=20ms TTL=61
Reply from 202.37.101.1: bytes=32 time=8ms TTL=61
Reply from 202.37.101.1: bytes=32 time=19ms TTL=61
Reply from 202.37.101.1: bytes=32 time=6ms TTL=61
Reply from 202.37.101.1: bytes=32 time=48ms TTL=61
Reply from 202.37.101.1: bytes=32 time=6ms TTL=61
Reply from 202.37.101.1: bytes=32 time=6ms TTL=61
Reply from 202.37.101.1: bytes=32 time=51ms TTL=61
Reply from 202.37.101.1: bytes=32 time=8ms TTL=61
Reply from 202.37.101.1: bytes=32 time=56ms TTL=61
Reply from 202.37.101.1: bytes=32 time=7ms TTL=61
Reply from 202.37.101.1: bytes=32 time=38ms TTL=61
Reply from 202.37.101.1: bytes=32 time=36ms TTL=61
Reply from 202.37.101.1: bytes=32 time=11ms TTL=61
Reply from 202.37.101.1: bytes=32 time=6ms TTL=61
Reply from 202.37.101.1: bytes=32 time=43ms TTL=61
Reply from 202.37.101.1: bytes=32 time=7ms TTL=61
Reply from 202.37.101.1: bytes=32 time=6ms TTL=61
Reply from 202.37.101.1: bytes=32 time=6ms TTL=61

Ping statistics for 202.37.101.1:
Packets: Sent = 19, Received = 19, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 6ms, Maximum = 56ms, Average = 20ms

I live about 1km away from Mt Eden/Dominion Road intersection; but I am on ADSL2, not VDSL.



1778 posts

Uber Geek
+1 received by user: 203


  Reply # 702687 17-Oct-2012 21:23 Send private message

@astrial: that looks a lot like I'm seeing - try it around 11:30pm \ 12pm (if you can, or maybe in the morning) I'm only seeing it in the evenings.

29 posts

Geek


  Reply # 702800 18-Oct-2012 10:34 Send private message

Yeah, my connection was poor the whole night last night, which was strange.  I'll do some more checking tonight.



Looks like I am connected to the Three Kings exchange (if I can read the Chorus network map correctly).

190 posts

Master Geek
+1 received by user: 1

Subscriber

  Reply # 703063 18-Oct-2012 16:43 Send private message

Hi all,

Can you post back with the following, trying to establish a pattern here:

What suburb/city you are in:
Which exchange, if you know:
What time the high latency starts:
What time the high latency ends:
Does your throughput drop while this is happening:
If throughput drops, to approx. what speed:

Thanks

29 posts

Geek


  Reply # 703228 18-Oct-2012 21:46 Send private message

Looks fine at this time, on this night (on the Three Kings exchange)

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=6ms TTL=61
Reply from 202.37.101.1: bytes=32 time=6ms TTL=61
Reply from 202.37.101.1: bytes=32 time=6ms TTL=61
Reply from 202.37.101.1: bytes=32 time=12ms TTL=61
Reply from 202.37.101.1: bytes=32 time=6ms TTL=61
Reply from 202.37.101.1: bytes=32 time=6ms TTL=61
Reply from 202.37.101.1: bytes=32 time=6ms 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=6ms TTL=61
Reply from 202.37.101.1: bytes=32 time=6ms TTL=61

Ping statistics for 202.37.101.1:
Packets: Sent = 12, Received = 12, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 5ms, Maximum = 12ms, Average = 6ms



1778 posts

Uber Geek
+1 received by user: 203


  Reply # 703241 18-Oct-2012 22:05 Send private message

@bender:

What suburb/city you are in: Mt Eden, Auckland
Which exchange, if you know: Mt Eden (Cabinet: MOD\AN)
What time the high latency starts: ~20:30
What time the high latency ends: ~23:30
Does your throughput drop while this is happening: Yes; the upload more noticeably but download on speedtests. 
If throughput drops, to approx. what speed: Not a huge amount according to speedtest; usually Speedtests are a very consistent 33/9.4  but during these higher latency periods it drops to around 28/7.  The speedtest graph also varies a lot more during these periods, whereas during the rest of the day it tends to be quite close to a straight line.  And whereas during the rest of the day it ramps pretty much immediately to the max throughput it takes a while in the evenings. 

 

EDIT: BTW this is what my latency currently looks like to trademe (just for something different)  I'll test again a bit later but I'm pretty sure it's usually around 5-6ms too or not far off:


C:\Users\Simon>ping trademe.co.nz /t

Pinging trademe.co.nz [202.162.72.2] with 32 bytes of data:
Request timed out.
Reply from 202.162.72.2: bytes=32 time=52ms TTL=250
Reply from 202.162.72.2: bytes=32 time=47ms TTL=250
Reply from 202.162.72.2: bytes=32 time=56ms TTL=250
Reply from 202.162.72.2: bytes=32 time=47ms TTL=250
Reply from 202.162.72.2: bytes=32 time=44ms TTL=250
Reply from 202.162.72.2: bytes=32 time=52ms TTL=250
Reply from 202.162.72.2: bytes=32 time=47ms TTL=250
Reply from 202.162.72.2: bytes=32 time=54ms TTL=250
Reply from 202.162.72.2: bytes=32 time=51ms TTL=250
Reply from 202.162.72.2: bytes=32 time=47ms TTL=250
Reply from 202.162.72.2: bytes=32 time=56ms TTL=250
Reply from 202.162.72.2: bytes=32 time=56ms TTL=250

Ping statistics for 202.162.72.2:
Packets: Sent = 13, Received = 12, Lost = 1 (7% loss),
Approximate round trip times in milli-seconds:
Minimum = 44ms, Maximum = 56ms, Average = 50ms





2975 posts

Uber Geek
+1 received by user: 174

Trusted
Subscriber

  Reply # 703246 18-Oct-2012 22:17 Send private message

Those cabinets have a 1gbps backhaul servicing around 200 customers. Is it really that likely for congestion ?!







1778 posts

Uber Geek
+1 received by user: 203


  Reply # 703248 18-Oct-2012 22:20 Send private message

Zeon: Those cabinets have a 1gbps backhaul servicing around 200 customers. Is it really that likely for congestion ?!


No idea, that's why I created the thread; What are the other possibilities?

EDIT: Not sure if they would be on the cabinet but I think the area on dominion road covers at least one internet cafe. 

1078 posts

Uber Geek
+1 received by user: 45


  Reply # 703257 18-Oct-2012 22:36 Send private message

sidefx:

C:\Users\Simon>ping trademe.co.nz /t

Pinging trademe.co.nz [202.162.72.2] with 32 bytes of data:


fwiw trademe is a bad place to ping to as it bounces between wellington and auckland randomly, snap's dns server is "anycast" which means it'll go to the closest location.   my trademe is currently resolving to 202.162.73.2 which pings consistently lower than 202.162.72.2



1778 posts

Uber Geek
+1 received by user: 203


  Reply # 703260 18-Oct-2012 22:39 Send private message

Hmm, good to know, that one is a bit lower:

Pinging 202.162.73.2 with 32 bytes of data:
Reply from 202.162.73.2: bytes=32 time=46ms TTL=250
Reply from 202.162.73.2: bytes=32 time=28ms TTL=250
Reply from 202.162.73.2: bytes=32 time=43ms TTL=250
Reply from 202.162.73.2: bytes=32 time=43ms TTL=250

Ping statistics for 202.162.73.2:
Packets: Sent = 4, Received = 4, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 28ms, Maximum = 46ms, Average = 40ms


Pinging 202.162.72.2 with 32 bytes of data:
Reply from 202.162.72.2: bytes=32 time=58ms TTL=250
Reply from 202.162.72.2: bytes=32 time=39ms TTL=250
Reply from 202.162.72.2: bytes=32 time=48ms TTL=250
Reply from 202.162.72.2: bytes=32 time=49ms TTL=250

Ping statistics for 202.162.72.2:
Packets: Sent = 4, Received = 4, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 39ms, Maximum = 58ms, Average = 48ms





EDIT: Of course they both look much better at 11:30 pm....

C:\Users\Simon>ping 202.162.73.2
Pinging 202.162.73.2 with 32 bytes of data:
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 = 4, Received = 4, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 6ms, Maximum = 6ms, Average = 6ms

C:\Users\Simon>ping 202.162.72.2

Pinging 202.162.72.2 with 32 bytes of data:
Reply from 202.162.72.2: bytes=32 time=14ms TTL=250
Reply from 202.162.72.2: bytes=32 time=14ms TTL=250
Reply from 202.162.72.2: bytes=32 time=14ms TTL=250
Reply from 202.162.72.2: bytes=32 time=14ms TTL=250

Ping statistics for 202.162.72.2:
Packets: Sent = 4, Received = 4, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 14ms, Maximum = 14ms, Average = 14ms

7682 posts

Uber Geek
+1 received by user: 266

Trusted
Subscriber

  Reply # 703605 19-Oct-2012 16:21 Send private message

Zeon: Those cabinets have a 1gbps backhaul servicing around 200 customers. Is it really that likely for congestion ?!


From memory the monthly fee for EUBA to Chorus gets you from your cabinet/exchange to the regional EAS (ethernet aggregation switch).

So for these guys in Mt Eden, 3 Kings that would be Mt Albert (I think).  

From there it would likely be EUBA tail extension (from Mt Albert to Mayoral Drive), then handover to Snap's Auckland network which is probably in/at Skytower.

They could be using alternative provider to backhaul from Mt Albert to their network (eg: Vector).

Suspect peak time congestion on Snap's transit from Snap's Auckland to the internet will be the issue, they've probably had significant customer growth and network improvements/upgrades haven't caught up.

Pure speculation though.



105 posts

Master Geek


  Reply # 703803 19-Oct-2012 22:20 Send private message

Ragnor:
Suspect peak time congestion on Snap's transit from Snap's Auckland to the internet will be the issue, they've probably had significant customer growth and network improvements/upgrades haven't caught up.

Pure speculation though.


This appears not to be the case as its not effecting everyone in Auckland simply a few isolated users - If this were the case you'd see it much more widespread.

Still more than likely a contention / degradation on the local cabinet. Regardless its best to email Snap followed by a phone call to make sure it gets to their more senior CSR's to take a look.

I do understand how hard this kind of fault is to work out where exactly the issue is! But its best to get all parties on-board as they will have different visability of the network.

My 2c.

The Game.
2897 posts

Uber Geek
+1 received by user: 468

Trusted
Think Concepts
Subscriber

  Reply # 703848 20-Oct-2012 00:09 Send private message

My pings are looking pretty good tonight:

mmurphy@router:~$ ping 202.37.101.1
PING 202.37.101.1 (202.37.101.1) 56(84) bytes of data.
64 bytes from 202.37.101.1: icmp_req=1 ttl=61 time=24.4 ms
64 bytes from 202.37.101.1: icmp_req=2 ttl=61 time=23.6 ms
64 bytes from 202.37.101.1: icmp_req=3 ttl=61 time=23.5 ms
64 bytes from 202.37.101.1: icmp_req=4 ttl=61 time=23.5 ms
64 bytes from 202.37.101.1: icmp_req=5 ttl=61 time=23.0 ms
64 bytes from 202.37.101.1: icmp_req=6 ttl=61 time=24.0 ms
64 bytes from 202.37.101.1: icmp_req=7 ttl=61 time=23.4 ms
^C
--- 202.37.101.1 ping statistics ---
7 packets transmitted, 7 received, 0% packet loss, time 6010ms
rtt min/avg/max/mdev = 23.049/23.672/24.413/0.454 ms

I have indeed noticed some degrade in speed / latency as of late, I thought it was my flatmates using the net getting the latest Ubuntu 12.10 distro but it turns out it wasn't - the Fritz!box is showing almost zero traffic going through.

Here are some graphs from my TrueNet probe:

 

Domestic Ping (Hourly Average - 1 week)



International Ping (Hourly Average - 1 week)




From what I can see here it seems that pinging Snap's own DNS server is showing delay during peak times, going overseas normally has no delay (assuming this is what the TrueNet probe is actually pinging)




Michael Murphy
[Twitter] [Last.fm] [IPv6 Sage]

Everything I say here is my own opinion and not that of my employer.



1778 posts

Uber Geek
+1 received by user: 203


  Reply # 703857 20-Oct-2012 00:34 Send private message

(Hopefully OK to share these) Mine are even less pretty, especially the throughput graphs. I assume because they test with fairly small files and with the higher potential throughput of VDSL it's more sensitive to increased latency? It looks like it's also affecting both domestic and international latency.

Domestic ping



International Ping


Download throughput


Upload Throughput

1 | 2 | 3 | 4 | 5
View this topic in a long page with up to 500 replies per page Create new topic

Twitter »
Follow us to receive Twitter updates when new discussions are posted in our forums:



Follow us to receive Twitter updates when news items and blogs are posted in our frontpage:



Follow us to receive Twitter updates when tech item prices are listed in our price comparison site:




News »

Trending now »
Hot discussions in our forums right now:

Does NZ need better gun laws?
Created by mattwnz, last reply by heylinb4nz on 2-Sep-2014 16:23 (104 replies)
Pages... 5 6 7


Warning: Rage Ahead - Campbell Live and childhood poverty
Created by kawaii, last reply by hangon on 2-Sep-2014 16:45 (88 replies)
Pages... 4 5 6


What tyre brand/model to look at ?
Created by Mark, last reply by Mark on 2-Sep-2014 15:14 (29 replies)
Pages... 2


VideoEZY OnDemand
Created by Andib, last reply by networkn on 2-Sep-2014 14:30 (57 replies)
Pages... 2 3 4


Judith Collins: I am resigning
Created by Presso, last reply by gzt on 2-Sep-2014 11:42 (109 replies)
Pages... 6 7 8


Cirque du Soleil Cellphone Hijack
Created by myopinion, last reply by PhantomNVD on 1-Sep-2014 18:01 (21 replies)
Pages... 2


Orcon Global Mode launched
Created by freitasm, last reply by shk292 on 1-Sep-2014 11:32 (132 replies)
Pages... 7 8 9


Lightbox press event release
Created by freitasm, last reply by IcI on 30-Aug-2014 17:54 (562 replies)
Pages... 36 37 38



Geekzone Live »
Try automatic live updates from Geekzone directly in your browser, without refreshing the page, with Geekzone Live now.

Are you subscribed to our RSS feed? You can download the latest headlines and summaries from our stories directly to your computer or smartphone by using a feed reader.

Alternatively, you can receive a daily email with Geekzone updates.