Congestion on the Auckland-Sydney route?

Forums › One New Zealand (including Vodafone, WxC, Farmside) › Congestion on the Auckland-Sydney route?

Laforet

22 posts

Geek

#205201 2-Nov-2016 17:18

Location: North Shore, Auckland

Package: Home UFB 30/10

Problem: For the last couple of days, connectivity has degraded significantly to be point of being unusable. Here is what the current situation looks like to me:

- NZ Servers: No issue at all, stable performance

- US West Coast: Sometimes erratic, but I could generally get 10-15 Mbps throughput and decent ping (~150ms) during non-peak hours.

- Australia: Normal ping (<30ms) within VFAU's network but download speed is throttled to about 3 MBps whereas upload speed appears to be uncapped. Once outside their own network, speed deteriorates to about 1 Mbps down and 3 Mbps up with very high lantency (150ms+ for hosts in Sydney).

- Asia: Pure horror show. Throughput to various providers with good interconnection at HKIX, SGIX or JPIX has been consistently below 1 Mbps with high packet loss and lantency spikes up to 400ms as long as they transit through Australia.

Anyone experiencing the same issue?

thecatsgoolies

414 posts

Ultimate Geek

Trusted

#1662900 2-Nov-2016 19:20

Trace routes / server names? no issue here on VodafoneNZ 100/20 UFB

frank6353

3 posts

Wannabe Geek

#1662906 2-Nov-2016 19:32

Similar thing is happening to my connection. Been experiencing packet loss while playing overwatch/ battlefield 1 (on AU server) for the past week or 2. My ping stays normal, only constant packet loss from 0-20%

(I'm with vodafone VDSL with 30m download, 5mb upload speed)

thecatsgoolies

414 posts

Ultimate Geek

Trusted

#1662914 2-Nov-2016 19:50

The new VodafoneNZ / Spark / Telstra Fibre Cable between NZ and OZ should add far more capacity when it goes live soon

Laforet

22 posts

Geek

#1662921 2-Nov-2016 20:02

thecatsgoolies:

Trace routes / server names? no issue here on VodafoneNZ 100/20 UFB

I've attached a partial trace to ato.gov.au and you can see the ping spike once it leaves vodafone's edge:

2    31 ms     4 ms     3 ms [redacted]
3     6 ms     3 ms     3 ms bvi-400.bgnzldv02.akl.vf.net.nz [203.109.180.242]
4     6 ms     3 ms     3 ms bvi-188.bgnzldv02.akl.vf.net.nz [203.109.180.241]
5    29 ms    27 ms    27 ms 10.123.80.9
6   157 ms   159 ms   155 ms ten-1-0-0-640.bdr03.syd03.nsw.VOCUS.net.au [114.31.200.221]
7   157 ms   155 ms   155 ms bundle-103.cor03.syd03.nsw.vocus.net.au [114.31.192.56]
8   157 ms   155 ms   155 ms ten-2-2.ces01.syd05.nsw.VOCUS.net.au [114.31.192.85]
9   157 ms   155 ms   159 ms 220.101.73.81

It has actually got somewhat better between the time of posting but they still need to fix their transit with Vocus - 150ms ping to Sydney is pretty bad by reasonable standards.

yitz

2077 posts

Uber Geek

#1662938 2-Nov-2016 20:22

Seems it might be coming back via an interconnect in the States.

route-views.optus.net.au>trace 118.92.0.1

Type escape sequence to abort.

Tracing the route to 118-92-0-1.dsl.dyn.ihug.co.nz (118.92.0.1)

1 vlan299.52gdc76ue2.optus.net.au (203.202.125.3) [AS 7474] 0 msec * 0 msec

2 220.101.73.82 [AS 7474] 0 msec 0 msec 4 msec

3 bundle-108.cor01.syd04.nsw.vocus.net.au (114.31.192.86) [AS 4826] [MPLS: Label 16305 Exp 0] 156 msec 204 msec 156 msec

4 ten-0-1-0-3-7.cor01.sjc01.ca.vocus.net (114.31.199.46) [AS 4826] [MPLS: Label 24015 Exp 0] 156 msec

ten-0-1-0-3-5.cor01.sjc01.ca.vocus.net (114.31.199.85) [AS 4826] [MPLS: Label 24015 Exp 0] 156 msec

ten-0-0-0-3-6.cor01.sjc01.ca.vocus.net (114.31.199.29) [AS 4826] [MPLS: Label 24015 Exp 0] 156 msec

5 bundle-100.bdr01.sjc01.ca.vocus.net (49.255.255.19) [AS 4826] 156 msec 160 msec 160 msec

6 as9500.cust.bdr03.sjc01.ca.VOCUS.net (114.31.199.138) [AS 4826] 160 msec 152 msec 156 msec

7 * * *

8 * * *

There could be many reasons for this, one of which is traffic is being scrubbed of DDoS attacks. This may explain the poor performance to international destinations.

Laforet

22 posts

Geek

#1663438 3-Nov-2016 18:47

yitz: Seems it might be coming back via an interconnect in the States. route-views.optus.net.au>trace 118.92.0.1 Type escape sequence to abort. Tracing the route to 118-92-0-1.dsl.dyn.ihug.co.nz (118.92.0.1) 1 vlan299.52gdc76ue2.optus.net.au (203.202.125.3) [AS 7474] 0 msec * 0 msec 2 220.101.73.82 [AS 7474] 0 msec 0 msec 4 msec 3 bundle-108.cor01.syd04.nsw.vocus.net.au (114.31.192.86) [AS 4826] [MPLS: Label 16305 Exp 0] 156 msec 204 msec 156 msec 4 ten-0-1-0-3-7.cor01.sjc01.ca.vocus.net (114.31.199.46) [AS 4826] [MPLS: Label 24015 Exp 0] 156 msec ten-0-1-0-3-5.cor01.sjc01.ca.vocus.net (114.31.199.85) [AS 4826] [MPLS: Label 24015 Exp 0] 156 msec ten-0-0-0-3-6.cor01.sjc01.ca.vocus.net (114.31.199.29) [AS 4826] [MPLS: Label 24015 Exp 0] 156 msec 5 bundle-100.bdr01.sjc01.ca.vocus.net (49.255.255.19) [AS 4826] 156 msec 160 msec 160 msec 6 as9500.cust.bdr03.sjc01.ca.VOCUS.net (114.31.199.138) [AS 4826] 160 msec 152 msec 156 msec 7 * * * 8 * * * There could be many reasons for this, one of which is traffic is being scrubbed of DDoS attacks. This may explain the poor performance to international destinations.

Thanks, it's back to normal this morning.

hio77

12999 posts

Uber Geek

ID Verified

Trusted

Lizard Networks

#1663439 3-Nov-2016 18:50

I've been doing some heavy overwatch play lately, Have not noticed any issues with congestion at all myself. solid 32ms throughout the whole play.

edit:

skiming over my smokepings i run, not noticeable problems at all which do not correlate with upstream congestion on my local side.

#include <std_disclaimer>

Any comments made are personal opinion and do not reflect directly on the position my current or past employers may have.