International Traffic issues on Orcon at moment?

Forums › 2degrees (including Slingshot, Orcon, Flip, Stuff Fibre, MyRepublic, 2talk and Vocus) › International Traffic issues on Orcon at moment?

1 | ... | 38 | 39 | 40 | 41 | 42 | 43 | 44 | 45 | 46 | 47 | 48 | ... | 49

231 posts

Master Geek

#886237 28-Aug-2013 22:07

Doing a tracert to us.blizzard.com, its taking a long time because of lots of "Request timed out." messages.

Microsoft Windows [Version 6.2.9200] (c) 2012 Microsoft Corporation. All rights reserved.
tracert us.blizzard.com
Tracing route to us.blizzard.com [12.129.242.31] over a maximum of 30 hops:
1 <1 ms 1 ms <1 ms BoB.Orcon [10.1.1.1] 2 15 ms 28 ms 12 ms euba.bng1.nct.orcon.net.nz [60.234.8.2] 3 25 ms 30 ms 50 ms xe-2-2-0.cre1.nct.orcon.net.nz [121.98.9.141] 4 16 ms 15 ms 19 ms 121.99.12.0 5 183 ms 188 ms 178 ms 38.122.92.162 6 176 ms 168 ms 164 ms te7-4.ccr01.sjc05.atlas.cogentco.com [38.122.92. 161] 7 171 ms 164 ms 166 ms te0-3-0-7.ccr22.sjc01.atlas.cogentco.com [154.54 .6.69] 8 181 ms 165 ms 163 ms te0-3-0-5.mpd22.sfo01.atlas.cogentco.com [154.54 .86.189] 9 177 ms 179 ms 187 ms te0-4-0-3.mpd22.mci01.atlas.cogentco.com [154.54 .45.85] 10 195 ms 206 ms 191 ms te0-0-0-5.ccr22.dfw01.atlas.cogentco.com [154.54 .5.157] 11 230 ms 227 ms 226 ms be2032.ccr21.dfw03.atlas.cogentco.com [154.54.6. 54] 12 203 ms 190 ms 191 ms 192.205.36.221 13 215 ms 214 ms 210 ms cr2.dlstx.ip.att.net [12.122.138.18] 14 202 ms 179 ms 185 ms cr2.la2ca.ip.att.net [12.122.28.178] 15 212 ms 207 ms 207 ms gar29.la2ca.ip.att.net [12.122.129.241] 16 * * * Request timed out. 17 227 ms 235 ms 222 ms mdf001c7613r0004-gig-10-1.lax1.attens.net [12.12 9.193.250] 18 * * * Request timed out. 19 * * * Request timed out. 20 * * * Request timed out. 21 * * * Request timed out. 22 * * * Request timed out. 23 * * * Request timed out. 24 * * * Request timed out. 25 * * * Request timed out. 26 * * * Request timed out. 27 * * * Request timed out. 28 * * * Request timed out. 29 * * * Request timed out. 30 * * * Request timed out.
Trace complete.

Will post tracert to my specific battlegroup in a moment.X

Sounddude

I fix stuff!
1928 posts

Uber Geek

Trusted

2degrees

Lifetime subscriber

#886241 28-Aug-2013 22:11

Rambaldi: Doing a tracert to us.blizzard.com, its taking a long time because of lots of "Request timed out." messages. Was going to tracert to my specific battlegroup right after but not sure how long it'll take as it is.

traceroute to us.blizzard.com (12.129.242.31), 64 hops max, 52 byte packets
1 bob.orcon (192.168.1.254) 1.894 ms 1.143 ms 1.066 ms
2 llu.bng1.nct.orcon.net.nz (60.234.8.1) 18.925 ms 17.700 ms 17.991 ms
3 xe-2-2-0.cre1.nct.orcon.net.nz (121.98.9.141) 20.839 ms 22.119 ms 17.244 ms
4 121.99.12.0 (121.99.12.0) 17.313 ms 39.355 ms 18.179 ms
5 38.122.92.162 (38.122.92.162) 145.024 ms 146.278 ms 146.400 ms
6 te7-4.ccr01.sjc05.atlas.cogentco.com (38.122.92.161) 145.033 ms 145.400 ms *
7 te0-3-0-7.ccr22.sjc01.atlas.cogentco.com (154.54.6.69) 148.665 ms    te0-2-0-4.ccr22.sjc01.atlas.cogentco.com (154.54.84.57) 147.081 ms    te0-2-0-4.ccr21.sjc01.atlas.cogentco.com (154.54.84.53) 172.411 ms
8 te0-1-0-5.ccr21.sfo01.atlas.cogentco.com (154.54.2.137) 173.046 ms    te0-2-0-4.mpd22.sfo01.atlas.cogentco.com (66.28.4.9) 173.076 ms    te0-1-0-5.ccr22.sfo01.atlas.cogentco.com (66.28.4.5) 150.887 ms
9 te0-2-0-2.mpd21.mci01.atlas.cogentco.com (154.54.6.166) 185.200 ms    te0-1-0-2.ccr22.mci01.atlas.cogentco.com (154.54.30.70) 185.187 ms    te0-3-0-1.ccr21.mci01.atlas.cogentco.com (154.54.24.110) 184.999 ms
10 te0-4-0-2.ccr21.dfw01.atlas.cogentco.com (154.54.46.202) 194.868 ms    te0-2-0-2.ccr22.dfw01.atlas.cogentco.com (154.54.25.209) 219.916 ms    te0-2-0-2.ccr21.dfw01.atlas.cogentco.com (154.54.5.221) 221.260 ms
11 be2032.ccr21.dfw03.atlas.cogentco.com (154.54.6.54) 221.696 ms 221.175 ms    be2031.ccr21.dfw03.atlas.cogentco.com (154.54.7.46) 441.382 ms
12 192.205.36.221 (192.205.36.221) 193.873 ms 195.494 ms 192.355 ms
13 cr2.dlstx.ip.att.net (12.122.138.18) 218.308 ms 218.008 ms 219.932 ms
14 cr2.la2ca.ip.att.net (12.122.28.178) 219.947 ms 218.506 ms 220.699 ms
15 gar29.la2ca.ip.att.net (12.122.129.241) 215.332 ms 216.950 ms 215.746 ms16 *

Looks OK to me.

Sounddude

I fix stuff!
1928 posts

Uber Geek

Trusted

2degrees

Lifetime subscriber

#886243 28-Aug-2013 22:12

Rambaldi: Doing a tracert to us.blizzard.com, its taking a long time because of lots of "Request timed out." messages.

Microsoft Windows [Version 6.2.9200] (c) 2012 Microsoft Corporation. All rights reserved.
tracert us.blizzard.com
Tracing route to us.blizzard.com [12.129.242.31] over a maximum of 30 hops:
1 <1 ms 1 ms <1 ms BoB.Orcon [10.1.1.1] 2 15 ms 28 ms 12 ms euba.bng1.nct.orcon.net.nz [60.234.8.2] 3 25 ms 30 ms 50 ms xe-2-2-0.cre1.nct.orcon.net.nz [121.98.9.141] 4 16 ms 15 ms 19 ms 121.99.12.0 5 183 ms 188 ms 178 ms 38.122.92.162 6 176 ms 168 ms 164 ms te7-4.ccr01.sjc05.atlas.cogentco.com [38.122.92. 161] 7 171 ms 164 ms 166 ms te0-3-0-7.ccr22.sjc01.atlas.cogentco.com [154.54 .6.69] 8 181 ms 165 ms 163 ms te0-3-0-5.mpd22.sfo01.atlas.cogentco.com [154.54 .86.189] 9 177 ms 179 ms 187 ms te0-4-0-3.mpd22.mci01.atlas.cogentco.com [154.54 .45.85] 10 195 ms 206 ms 191 ms te0-0-0-5.ccr22.dfw01.atlas.cogentco.com [154.54 .5.157] 11 230 ms 227 ms 226 ms be2032.ccr21.dfw03.atlas.cogentco.com [154.54.6. 54] 12 203 ms 190 ms 191 ms 192.205.36.221 13 215 ms 214 ms 210 ms cr2.dlstx.ip.att.net [12.122.138.18] 14 202 ms 179 ms 185 ms cr2.la2ca.ip.att.net [12.122.28.178] 15 212 ms 207 ms 207 ms gar29.la2ca.ip.att.net [12.122.129.241] 16 * * * Request timed out. 17 227 ms 235 ms 222 ms mdf001c7613r0004-gig-10-1.lax1.attens.net [12.12 9.193.250] 18 * * * Request timed out. 19 * * * Request timed out. 20 * * * Request timed out. 21 * * * Request timed out. 22 * * * Request timed out. 23 * * * Request timed out. 24 * * * Request timed out. 25 * * * Request timed out. 26 * * * Request timed out. 27 * * * Request timed out. 28 * * * Request timed out. 29 * * * Request timed out. 30 * * * Request timed out.
Trace complete.

Will post tracert to my specific battlegroup in a moment.X

Yes that is expected. Blizzard drop ICMP at its border, which shows timeouts in a traceroute.

Rambaldi

231 posts

Master Geek

#886245 28-Aug-2013 22:16

Ahh okay, its basically same results to my specific battlegroup. Myself and the other Orcon customer are the only two people unable to raid tonight, we always know to message each other when one of us is lagging, and we always find all the other NZer's in our guild are fine, and we aren't.

mercutio

1392 posts

Uber Geek

#886247 28-Aug-2013 22:16

Sounddude:
Rambaldi: Doing a tracert to us.blizzard.com, its taking a long time because of lots of "Request timed out." messages. Was going to tracert to my specific battlegroup right after but not sure how long it'll take as it is.

traceroute to us.blizzard.com (12.129.242.31), 64 hops max, 52 byte packets
1 bob.orcon (192.168.1.254) 1.894 ms 1.143 ms 1.066 ms
2 llu.bng1.nct.orcon.net.nz (60.234.8.1) 18.925 ms 17.700 ms 17.991 ms
3 xe-2-2-0.cre1.nct.orcon.net.nz (121.98.9.141) 20.839 ms 22.119 ms 17.244 ms
4 121.99.12.0 (121.99.12.0) 17.313 ms 39.355 ms 18.179 ms
5 38.122.92.162 (38.122.92.162) 145.024 ms 146.278 ms 146.400 ms
6 te7-4.ccr01.sjc05.atlas.cogentco.com (38.122.92.161) 145.033 ms 145.400 ms *
7 te0-3-0-7.ccr22.sjc01.atlas.cogentco.com (154.54.6.69) 148.665 ms    te0-2-0-4.ccr22.sjc01.atlas.cogentco.com (154.54.84.57) 147.081 ms    te0-2-0-4.ccr21.sjc01.atlas.cogentco.com (154.54.84.53) 172.411 ms
8 te0-1-0-5.ccr21.sfo01.atlas.cogentco.com (154.54.2.137) 173.046 ms    te0-2-0-4.mpd22.sfo01.atlas.cogentco.com (66.28.4.9) 173.076 ms    te0-1-0-5.ccr22.sfo01.atlas.cogentco.com (66.28.4.5) 150.887 ms
9 te0-2-0-2.mpd21.mci01.atlas.cogentco.com (154.54.6.166) 185.200 ms    te0-1-0-2.ccr22.mci01.atlas.cogentco.com (154.54.30.70) 185.187 ms    te0-3-0-1.ccr21.mci01.atlas.cogentco.com (154.54.24.110) 184.999 ms
10 te0-4-0-2.ccr21.dfw01.atlas.cogentco.com (154.54.46.202) 194.868 ms    te0-2-0-2.ccr22.dfw01.atlas.cogentco.com (154.54.25.209) 219.916 ms    te0-2-0-2.ccr21.dfw01.atlas.cogentco.com (154.54.5.221) 221.260 ms
11 be2032.ccr21.dfw03.atlas.cogentco.com (154.54.6.54) 221.696 ms 221.175 ms    be2031.ccr21.dfw03.atlas.cogentco.com (154.54.7.46) 441.382 ms
12 192.205.36.221 (192.205.36.221) 193.873 ms 195.494 ms 192.355 ms
13 cr2.dlstx.ip.att.net (12.122.138.18) 218.308 ms 218.008 ms 219.932 ms
14 cr2.la2ca.ip.att.net (12.122.28.178) 219.947 ms 218.506 ms 220.699 ms
15 gar29.la2ca.ip.att.net (12.122.129.241) 215.332 ms 216.950 ms 215.746 ms16 *

Looks OK to me.

the traceroute isn't as bad as the packet loss to that host, which of course is hard to monitor because they filter icmp, but if you do a tcpdump you'll see a lot of retransmissions. it's cogent->at&t being saturated, and it seems cogent are now sending to at&t much further out. if at all possible to change at&t bgp path to not go out via cogent would probably be a lot better. (battle net is on another AS after some of the other traffic, but it's still single homed at&t.. )

on that note, it seems cogent are up to having peering issues with verizon, at&t, and comcast now. all simultaneously. afaik it's mostly about cogent hosting netflix.

the cogent->at&t dispute is pretty big, it seems they're congesting in los angeles, and chicago, and this path looks like dallas...

Rambaldi

231 posts

Master Geek

#886249 28-Aug-2013 22:19

Yeah when I say its Orcon playing up, I actually just mean its their routing partners/methods, as the other NZ players in my guild almost never have the same issues as myself and the other Orcon customer.

mercutio

1392 posts

Uber Geek

#886250 28-Aug-2013 22:19

Rambaldi: Yeah when I say its Orcon playing up, I actually just mean its their routing partners/methods, as the other NZ players in my guild almost never have the same issues as myself and the other Orcon customer.

well it's mostly a cogent issue, luckily it's outbound, so they should be able to route all at&t traffic not via cogent, if they're willing to adjust their routing so.

the cogent/at&t thing isn't really orcon's fault. it's kind of political, with cogent not wanting to pay money to at&t and at&t not wanting to pay money to cogent. but in this instance i want to blame blizzard for single homing to at&t. (using only one upstream provider) orcon have level3, cogent, telstraclear, and telecom afaik. routing outbound via telecom or level3 to at&t should dramatically improve. unfortunately, at&t are apparently sending traffic from themselves out via cogent when people have cogent and other providers, .. but most of the congestion is sending not receiving, so it'd prob make it 140 to 240 msec, rather than 240 to 2000+ msec.

one workaround is to use a vpn in los angeles that doesn't have cogent as an upstream at all.

Samsung

Shop now on Samsung phones, tablets, TVs and more (affiliate link).

Sounddude

I fix stuff!
1928 posts

Uber Geek

Trusted

2degrees

Lifetime subscriber

#886252 28-Aug-2013 22:23

mercutio:
the traceroute isn't as bad as the packet loss to that host, which of course is hard to monitor because they filter icmp, but if you do a tcpdump you'll see a lot of retransmissions. it's cogent->at&t being saturated, and it seems cogent are now sending to at&t much further out. if at all possible to change at&t bgp path to not go out via cogent would probably be a lot better. (battle net is on another AS after some of the other traffic, but it's still single homed at&t.. )

on that note, it seems cogent are up to having peering issues with verizon, at&t, and comcast now. all simultaneously. afaik it's mostly about cogent hosting netflix.

the cogent->at&t dispute is pretty big, it seems they're congesting in los angeles, and chicago, and this path looks like dallas...

Depressing isn't it.

Will have a play in the morning, see if I can get the AT&T traffic to go via Level3.

Rambaldi

231 posts

Master Geek

#886258 28-Aug-2013 22:29

It wasn't like this during the day... not that I was on wow the entire time, normally anytime something like this happens it's fixed relatively quickly. Although with the current content in wow I doubt there will be a raid 2moro night as there won't be anything left to kill, oh well. Still completely unsure why my gmail & gmail chat keeps dropping though, just weird, maybe they use similar routing I dunno.

Sounddude

I fix stuff!
1928 posts

Uber Geek

Trusted

2degrees

Lifetime subscriber

#886259 28-Aug-2013 22:30

Rambaldi: It wasn't like this during the day... not that I was on wow the entire time, normally anytime something like this happens it's fixed relatively quickly. Although with the current content in wow I doubt there will be a raid 2moro night as there won't be anything left to kill, oh well. Still completely unsure why my gmail & gmail chat keeps dropping though, just weird, maybe they use similar routing I dunno.

Could be all related and you have another issue entirely.

mercutio

1392 posts

Uber Geek

#886260 28-Aug-2013 22:32

Rambaldi: It wasn't like this during the day... not that I was on wow the entire time, normally anytime something like this happens it's fixed relatively quickly. Although with the current content in wow I doubt there will be a raid 2moro night as there won't be anything left to kill, oh well. Still completely unsure why my gmail & gmail chat keeps dropping though, just weird, maybe they use similar routing I dunno.

the issues happen primarily in US peak time, .. it's probably fine at this time. google have completely different routing.

Scotsman

83 posts

Master Geek

#886600 29-Aug-2013 15:45

With regards to the nightly plummet in speeds to Australia in particular, will Orcon be making any kind of 'official' statement to its customers regarding the plan to rectify this.

It seems we are the only ISP customers to get this daily, should I say nightly drop in service. With speeds losing up to 80% it would be nice of Orcon to actually put out some clear communication on exactly when and how this will be resolved.

Ive been moved to a new exchange only to find that my connection is great during the day but shocking at night, whichis when most customers need their connection.

Clearly this is not the service we are supposed to be paying for.

TechSol

298 posts

Ultimate Geek

Technical Solutions Aust

#886624 29-Aug-2013 16:22

wow, this still happens? I remember starting a thread about this sort of thing back in 2009, and I have seen several since.

Same problems, different day. Vote with your feet folks.

Rambaldi

231 posts

Master Geek

#886650 29-Aug-2013 17:31

In the past it's been that Orcon runs out of International Bandwidth, so they slow everything down for a few weeks while they pretend they haven't run out. Then they order more then everything is fine at night again for a good few months but hardly ever admit to doing anything to fix the problem and act like we were all imaging it. It's happened in a circle so many times that I normally don't bother saying anything, I just wait and it gets fixed. Only reason I stay with Orcon is I download a large amount and other ISP's in NZ are filthy scoudrels when it comes to data caps. Still paying 220ish a month on the 1TB plan that Orcon no longer offer, because I don't like how shady the all you can eat plan sounds in the fine print, with the whole we can charge you more if you download too much, without saying how much too much is.

Scotsman

83 posts

Master Geek

#886655 29-Aug-2013 17:41

This connection was around 12 Meg down around 3pm today....

1 | ... | 38 | 39 | 40 | 41 | 42 | 43 | 44 | 45 | 46 | 47 | 48 | ... | 49