[TelstraClear] TelstraClear Cable connections appear to be rate shaped in one way or another

Forums › One New Zealand (including Vodafone, WxC, Farmside) › [TelstraClear] TelstraClear Cable connections appear to be rate shaped in one way or another

1 | ... | 6 | 7 | 8 | 9 | 10 | 11 | 12

176 posts

Master Geek

#540718 3-Nov-2011 13:50

Ragnor: It seems to be Reach or upstream rather than a Telstraclear transparent proxy issue.

Lucky no one is holding their breath.

It was never a Transproxy issue, my traffic was sFTP over SSH (port 22 and port 2222 - neither made a difference) - end to end encrypted, not proxyable.

Incidentally, last they contacted me, TelstraClear were trying to set up a test bench off of a cable modem in Wellington, because they could not replicate the problem properly elsewhere. That suggests to me the problem may be closer to home than Reach.

I have no idea if they did set up their test bench, or if they conducted further tests, however, the problem remains, and may even be a little worse, so clearly nothing has been done about it.

jpollock

600 posts

Ultimate Geek

Trusted

#548240 22-Nov-2011 09:00

Just letting everyone know that we haven't forgotten... Any updates?

---
http://blog.jason.pollock.ca

jnawk

176 posts

Master Geek

#548272 22-Nov-2011 10:01

jpollock: Just letting everyone know that we haven't forgotten... ?Any updates?

Not really. The speeds are up and down. Sometimes good, sometimes not good. Which, I guess, is better than them being always not good.

As for official updates, from TC? Not a peep. I keep prodding Gary now and then, but nothing other than "I'll take a look" ever comes out of him.

DonGould

3892 posts

Uber Geek

#549855 25-Nov-2011 11:22

jnawk:
jpollock: Just letting everyone know that we haven't forgotten... ?Any updates?

Not really. The speeds are up and down. Sometimes good, sometimes not good. Which, I guess, is better than them being always not good.

As for official updates, from TC? Not a peep. I keep prodding Gary now and then, but nothing other than "I'll take a look" ever comes out of him.

jpollock, I've been following this issue as well.

My current take is that the peering that Telstra have just isn't as good as the peering Xnet have been able to secure.

Peering is a black art in relationship management and when providers/bean counters annoy people to much then you can see how this goes pear share fast. Just look at what happened to our Google feed as an example.

Talking with a contact in .au, I've learnt that we're not the only customers to be impacted who are traveling the Reach network.

I suspect that these issues are impacting why Telstra aren't giving us the 100mbit fruit. I have no clear evidence to prove any theory over the other, but anyone with have a clue can see that if customers start seeing that they can only pull 2mbit over a 100mbit tail, then things will go pear shape even more from a customer perception point of view.

It's on my radar to bounce this issue back to .au to have a closer look, but I've currently got a big list in there and I really don't think that this issue is priority over just getting some of the current issues addressed.

Personally I'd like to see the gap between .au and .nz data prices sorted over speed to .us/.eu

16 times the cost is just silly.

Promote New Zealand - Get yourself a .kiwi.nz domain name!!!

Check out mine - i.am.a.can.do.kiwi.nz - don@i.am.a.can.do.kiwi.nz

jnawk

176 posts

Master Geek

#583456 19-Feb-2012 15:40

No doubt this thread probably bores people now, but things haven't really changed. I notice I can get full speed if I direct my traffic across my IPv6 tunnel, terminated by ACSdata in Wellington.

jpollock

600 posts

Ultimate Geek

Trusted

#583537 19-Feb-2012 19:13

I'm still watching, I just figure that TCL's not going to do anything about it.

---
http://blog.jason.pollock.ca

Ragnor

8219 posts

Uber Geek

Trusted

#583663 20-Feb-2012 02:40

Considered changing ISP?

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.

jpollock

600 posts

Ultimate Geek

Trusted

#583686 20-Feb-2012 08:02

Ragnor: Considered changing ISP?

Not really. My friends' experience on DSL is pretty poor, and when I need actual bandwidth, I'll set up a VPN that uses UDP, hopefully avoiding the per-socket limitation entirely. If that doesn't work, then I'll go to load balancing across multiple VPN connections. 7-8 should get me the full 15mbps.

Rate limiting per socket will only slow down consumers for so long, it only takes one person to write up and package how to load balance across multiple VPN connections.

Heck, there might even be a market for a NZ VPN provider with a peering agreement with TCL/TNZ/etc...

---
http://blog.jason.pollock.ca

mercutio

1392 posts

Uber Geek

#583850 20-Feb-2012 12:59

jpollock:
Ragnor: Considered changing ISP?

Not really. My friends' experience on DSL is pretty poor, and when I need actual bandwidth, I'll set up a VPN that uses UDP, hopefully avoiding the per-socket limitation entirely. If that doesn't work, then I'll go to load balancing across multiple VPN connections. 7-8 should get me the full 15mbps.

Rate limiting per socket will only slow down consumers for so long, it only takes one person to write up and package how to load balance across multiple VPN connections.

Heck, there might even be a market for a NZ VPN provider with a peering agreement with TCL/TNZ/etc...

have you tried a download accelerator? axel is an example for linux. not sure what is out for windows.

i think you'll find that it's consistent low grade packet loss causing issues.

the easiest solution is to bounce your traffic through a server in los angeles to your remote destination. with tcp. you can setup something like redir with a firewall only allowing connections to your ip...

not sure if there is an easy way to do it transparently with vpn. but i think you'll find vpn on it's own doesn't help.

jpollock

600 posts

Ultimate Geek

Trusted

#584187 20-Feb-2012 21:24

mercutio:

have you tried a download accelerator? axel is an example for linux. not sure what is out for windows.

A wget replacement won't fix the problem for other TCP connections such as video streaming. :)

mercutio:

i think you'll find that it's consistent low grade packet loss causing issues.

Yes, the impressive thing is that it doesn't affect UDP traffic, and limits sockets to the rate they would achieve at that ping without the large window extension. That indicates that it is intentional behaviour of the network. We've diagnosed it as a problem with Reach, Telstra Clear's upstream provider. :)

mercutio:

the easiest solution is to bounce your traffic through a server in los angeles to your remote destination. with tcp. you can setup something like redir with a firewall only allowing connections to your ip...

not sure if there is an easy way to do it transparently with vpn. but i think you'll find vpn on it's own doesn't help.

Well, since LA is the wrong side of Reach, we still won't see any performance benefits, which is why you want a socket to a VPN provider here in NZ that uses a different international traffic provider, or else a VPN that uses UDP or load balanced TCP. :) You could even implement some of Google's proposed TCP changes (quick start/etc) to reduce the number of round trips to the remote VPN endpoint. :)

I'd like my 10mbps per socket please!

Jason

---
http://blog.jason.pollock.ca

Talkiet

4792 posts

Uber Geek

Trusted

#584215 20-Feb-2012 22:18

jpollock:
[snip]That indicates that it is intentional behaviour of the network. [snip]

I know you're still upset, but I think this inference you're making is illogical, unlikely, and wrong.

Cheers - N

Please note all comments are from my own brain and don't necessarily represent the position or opinions of my employer, previous employers, colleagues, friends or pets.

jpollock

600 posts

Ultimate Geek

Trusted

#584248 20-Feb-2012 22:53

How so? To recap:

We've got data showing UDP traffic working at full line rate, the TCP large window extensions supported by both endpoints and passed unchanged. We've got a different NZ ISP exhibiting twice the per-socket performance. We've got TCL admitting there's a problem and admitting that it's a problem with their upstream provider. However, there's been no change, which indicates that it is designed in.

There are many valid reasons to rate limit using packet loss, for one thing it's going to be hidden from the tools that look for RSTs/etc. The simplest "valid" reason is that it's an easy way to inject some "fairness" into streaming video downloads.

It doesn't have to be real, designed in "drop the packet to limit the socket". It could just be a buffer per socket which is used to forward packets down multiple links which then overflows. If the buffer is chosen by source/dest IP+port, that would have the same effect. So, it's congestion, and TCP does what it should do, but it's not real congestion - I can open a second socket and get the same traffic again, all the way up to a full 15mbps.

---
http://blog.jason.pollock.ca

jnawk

176 posts

Master Geek

#584250 20-Feb-2012 22:55

Talkiet:
jpollock:
[snip]That indicates that it is intentional behaviour of the network. [snip]

I know you're still upset, but I think this inference you're making is illogical, unlikely, and wrong.

Cheers - N

perhaps you might like to re-review the 11 prior pages of this thread - its all been thrashed to death, and the conclusion is very much that it is a network issue. Telstra have even admitted as much to me.

But they never solved it. others have insinuated that i have shown their peerimg arrangememts to be poor and that they should be embarassed, but the nuts and bolts of it is they never solved it, and i eventually lost interest, found a work around that works to one degree or other (its a pretty useless solution, i would rather they sort their crap out) and gave up banging away about it to gary, who had been getting slacker and slacker at replying to my email.

Talkiet

4792 posts

Uber Geek

Trusted

#584253 20-Feb-2012 22:59

jpollock: How so? To recap:
[snip]We've got TCL admitting there's a problem and admitting that it's a problem with their upstream provider. However, there's been no change, which indicates that it is designed in.
[snip]

Here's where I disagree... Failing to completely optimise something doesn't mean they designed precisely the behaviour you are apparently observing into the network.

I'm not disputing it's interesting from an academic standpoint, it's just I don't believe that they have designed this behaviour into the network. Mind you, you've taken some very interesting interpretations elsewhere in this thread so it's entirely possible you believe that a company using (for example) default settings out of the box means they deliberately and precisely understand and fully endorse the resulting behaviour, even if in fact the service is entirely fit for purpose and the cost of modifying it would far exceed any benefit it would bring them.

Cheers - N

Please note all comments are from my own brain and don't necessarily represent the position or opinions of my employer, previous employers, colleagues, friends or pets.

jnawk

176 posts

Master Geek

#584259 20-Feb-2012 23:08

Talkiet: Mind you, you've taken some very interesting interpretations elsewhere in this thread so it's entirely possible you believe that a company using (for example) default settings out of the box means they deliberately and precisely understand and fully endorse the resulting behaviour, even if in fact the service is entirely fit for purpose and the cost of modifying it would far exceed any benefit it would bring them.

Cheers - N

Mate, lets not start a pissing match, but if you believe for one moment that telstra don't have capable people who understand their stuff, then you do them a disservice.

1 | ... | 6 | 7 | 8 | 9 | 10 | 11 | 12