332 replies to this topic

[N0MAD]

I know that on these forums people were suggesting changing your Proxy to improve pings, have any of you at some stage changed your proxy? because thats one of the ways to force your route, and you are somehow forcing it because you are routing to a VPS, notice the lines i pointed out those are definately VPSs.

[ShinVector]

N0MAD, on 02 September 2014 - 10:43 PM, said:

Telstra cant do anything about people forcing their routs thru self appointed proxy's.

Hmmmm... Do you believe you understand how routing works ?

Regards,
Network Engineer

Edit:
Are people getting VPS (Virtual Private Server) and VPN (Virtual Private Network) confused ?

Edited by ShinVector, 02 September 2014 - 11:13 PM.

[John1352]

AUSwarrior24, on 02 September 2014 - 11:03 PM, said:

Calling my ISP is going to be the equivalent of poking a turtle with a stick- they have shocking customer service.

Biggest understatement ever. It's like poking Uluru with a blade of grass.

[Egomane]

Nauht, on 02 September 2014 - 10:55 PM, said:

So it is pure coincidence then? Now I'm no network guy but how come my pings to other NA games have't been affected.. I play ESO, SWOTR, DCS, SC and some other casual stuff.

And so if the data centre hadn't moved then I would be getting this ping increase regardless.. just for MWO?

Coincidences do happen but not when it's this specific.

Ok, as a network guy I can tell you this... (simplified version)

A router has a routing table. This routing table is telling the router where it needs to send a packet, if it needs to reach a certain adress.
If the routing table is telling the router ABC it needs to send a packet for address w.x.y.z through router BCA it will use that route.
If address w.y.x.z is geographically right next to w.x.y.z, but the routing table is telling the router to use a different route to it, for example through router CAB, then router ABC will do so, as the router is stupid and does not know geography.

So in other words... yes, this is a coincidence.

[Impyrium]

Egomane, on 02 September 2014 - 11:15 PM, said:

So in other words... yes, this is a coincidence.

The question is, at least for me, how and if this is fixable, do we do something about it, or the ISP?

[Oderint dum Metuant]

N0MAD, on 02 September 2014 - 11:05 PM, said:

I know that on these forums people were suggesting changing your Proxy to improve pings, have any of you at some stage changed your proxy? because thats one of the ways to force your route, and you are somehow forcing it because you are routing to a VPS, notice the lines i pointed out those are definately VPSs.

Your getting yourself confused; it has nothing to do with VPS or VPN.

It's a plain poor routing issue; theres a chap on another ISP in aus who routes correctly inside 8 hops without going through Europe in another thread.

Most Telstra customers are being routed badly that's all there is too it

AUSwarrior24, on 02 September 2014 - 11:20 PM, said:

The question is, at least for me, how and if this is fixable, do we do something about it, or the ISP?

You guys need to band together and get Telstra to notice it.
If they have a forum manned by CSRs this will usually do the trick; show them the tracers and question the routing through Europe.

[Nauht]

Egomane, on 02 September 2014 - 11:15 PM, said:

Ok, as a network guy I can tell you this... (simplified version)

A router has a routing table. This routing table is telling the router where it needs to send a packet, if it needs to reach a certain adress.
If the routing table is telling the router ABC it needs to send a packet for address w.x.y.z through router BCA it will use that route.
If address w.y.x.z is geographically right next to w.x.y.z, but the routing table is telling the router to use a different route to it, for example through router CAB, then router ABC will do so, as the router is stupid and does not know geography.

So in other words... yes, this is a coincidence.

Thanks, I get it now. So it doesn't matter whether the game server was in the US or not... routers only look use their routing table to an address, and each address will be different.

So wouldn't the move have changed the address at PGI's end? Thus creating a new routing table for it? Cos my ping was acceptable before the move, supposedly on the old address.

If that's the case then, yeah, I see that there would be nothing that PGI can really do (talking with no tech experience) as it's like the postman still taking the mail the long way to an address. You can only tell the postman where to go.

But it still ties in with the data centre change doesn't it? You're changing the actual address then.

[N0MAD]

ShinVector, on 02 September 2014 - 11:11 PM, said:

Hmmmm... Do you believe you understand how routing works ?

Regards,
Network Engineer

Edit:
Are people getting VPS (Virtual Private Server) and VPN (Virtual Private Network) confused ?

So Mr Network Engineer pls explain how their route is being routed thru a VPS? because if you look at lines 11 and 13 of the two traces i pointed out those addies are VPS.
Ego, routes can be forced right? by setting proxy, is that not correct?
So the routing table routes you thru your network, but its hubs that route you thru the internet finding the next closest or direct hub to your destination, is that fairly correct? but you can change how this works by setting your own proxy?
Now i know this because you can use a prog like, Chris pc proxy, i can set a proxy say in Germany and then that forces my routing thru germany, deviating from what would be the normal rout my connection would take.

Edited by N0MAD, 02 September 2014 - 11:42 PM.

[Egomane]

@N0MAD
You can manipulate the routing by using VPNs or proxy servers. That is correct. Your provider router will then route you to the target adress of either of those and you will be routed from there to the final target destination.

@Naught
PGI did not need to get a new IP address, for this routing change to happen. A relocation behind a different endpoint router can have the same effect.
To extend your postman analogy:
- You have a mailbox (IP address) in 221b Baker Street, London (physical Address).
- You always got your mail (data) in 9 a.m.
- You relocate to 223 Baker Street, London
- You removed your old mailbox in 221b and mounted it on 223
- For some reason another postman (router) is responsible for 223 Baker Street, though it's very close to 221b
- This new postman takes another route and will deliver your post at 11 a.m. as his route is longer then that of the old one.
- The only thing you can do, to try to get your mail again at 9 a.m. is to contact the mail delivery service (provider) and ask them to change the route your new postman is using.

So, yes... it is related to the datacenter change. But it is beyond the technical responsibility of PGI. They can't fix the routing of your ISP and they couldn't anticipate it.

Edited by Egomane, 02 September 2014 - 11:56 PM.

[ShinVector]

N0MAD, on 02 September 2014 - 11:34 PM, said:

So Mr Network Engineer pls explain how their route is being routed thru a VPS? because if you look at lines 11 and 13 of the two traces i pointed out those addies are VPS.
Ego, routes can be forced right? by setting proxy, is that not correct?
So the routing table routes you thru your network, but its hubs that route you thru the internet finding the next closest or direct hub to your destination, is that fairly correct? but you can change how this works by setting your own proxy?

Firstly please explain what you are referring to as VPS ? Lost in translation to me. VPS has nothing to do with routing at all. Routers and switches yes !
By hop 11 to 13 you mean the traceroute from Naduk... ?

Basically how the router goes.
1. Telstra
2. HOP10 - AS3320 Deutsche Telekom AG
3. HOP11 - AS16276 OVH SAS

http://bgp.he.net/ip/217.239.40.230
http://bgp.he.net/ip/94.23.122.153

Routing is largely dependant on the resources the ISP, their routing policy has and the peers they have.
Often this not free.. Eg. Like in my country where you have to pay for your upstream links.

The only immediate workaround is to try using VPNs. Vypr VPN, Astrill VPN.
The only gaming proxy I know WTfast does not work with MWO.

Else you ISP has to find a better route to pgi's new DC routers.
Might need PGI ISP to do something as well.

Edited by ShinVector, 03 September 2014 - 12:03 AM.

[Rashkae]

The biggest help would be for PGI to get the servers out of Canada.

The hop from NA to Canada adds 120ms to my Ping.

[N0MAD]

ShinVector, on 02 September 2014 - 11:57 PM, said:

Firstly please explain what you are referring to as VPS ? Lost in translation to me. VPS has nothing to do with routing at all. Routers and switches yes !
By hop 11 to 13 you mean the traceroute from Naduk... ?

Basically how the router goes.
1. Telstra
2. HOP10 - AS3320 Deutsche Telekom AG
3. HOP11 - AS16276 OVH SAS

http://bgp.he.net/ip/217.239.40.230
http://bgp.he.net/ip/94.23.122.153

Routing is largely dependant on the resources the ISP, their routing policy has and the peers they have.
Often this not free.. Eg. Like in my country where you have to pay for your upstream links.

The only immediate workaround is to try using VPNs. Vypr VPN, Astrill VPN.
The only gaming proxy I know WTfast does not work with MWO.

Else you ISP has to find a better route to pgi's new DC routers.
Might need PGI ISP to do something as well.

Naduk, on 02 September 2014 - 06:30 PM, said:

no its not Nauht, you can even check where your net traffic is going
open a cmd window and type tracert then web address

as you can see here my traffic is going through France to get to the USA

AUSwarrior24, on 02 September 2014 - 10:47 PM, said:

Hm. Mine looks pretty identical to yours.



To be honest, knowing Telstra, they're not going to give a flying rat's ass if a few people are having trouble like this. They follow the 'minimum viable product' ideology.

Ok the two traces i was refering to.
The first trace, hop 11 is a VPS, trace two hop 13 is to the same VPS, so those two routes seem to have been forced to that VPS, i imagine thats why they are going the Asian/Euro rout rather than the normal USA/Canada route.
Youre ISP routes you to the Telstra international hub, the hub takes over and finds the most direct rout to the next hub on the rout to your destination, unless of course the route is forced, then your ISP takes you to the international hub and then its forced to go to a directed proxy, is that not correct? Ego may want to comment as he is in the bizz.
So seeing that both those routes are going to a PRIVATE server (VPS) i would guess they are being forced that way.
I cant see an Aussie ISP directing their traffic thru a VPS in Germany hosting a French server..

[Oderint dum Metuant]

Except it's not just those two peoples routes......there's a lot more they haven't all set up the same private server to route too and it only began post server move.....

Your barking up the wrong tree nomad; there are smaller aus ISPs routing the direct way.

Edited by DV McKenna, 03 September 2014 - 01:37 AM.

[Kamikaze Viking]

Ok then. Here's my trace. From Adelaide on Internode (although it seems they've renamed the infrastructure since the iinet merger)

You can see it goes from Adelaide to Sydney, then California then Canada.

I'm guessing its a Telstra Problem (as always)

C:\Windows\system32>Tracert mwomercs.com

Tracing route to mwomercs.com [192.99.109.192]
over a maximum of 30 hops:

1 2 ms 1 ms 1 ms 192.168.1.1
2 20 ms 20 ms 20 ms lns21.adl2.on.ii.net [203.16.215.199]
3 20 ms 20 ms 22 ms xe10-0-4.cr1.adl2.on.ii.net [150.101.33.203]
4 187 ms 186 ms 186 ms ae0.cr1.adl6.on.ii.net [150.101.33.3]
5 40 ms 40 ms 46 ms ae4.br1.syd7.on.ii.net [150.101.33.34]
6 187 ms 186 ms 186 ms te0-1-1-2.br1.sjc2.on.ii.net [150.101.33.193]
7 219 ms 222 ms 340 ms eqx.snj.ovh.net [206.223.116.193]
8 190 ms 187 ms 189 ms 198.27.73.249
9 * * 240 ms 198.27.73.184
10 * 304 ms 255 ms 198.27.73.179
11 257 ms 257 ms 257 ms bhs-g2-6k.qc.ca [198.27.73.7]
12 253 ms 253 ms 253 ms 198.27.73.233
13 256 ms 256 ms 256 ms mwomercs.com [192.99.109.192]

Trace complete.

Edited by Kamikaze Viking, 03 September 2014 - 01:42 AM.

[N0MAD]

DV McKenna, on 03 September 2014 - 01:26 AM, said:

Except it's not just those two peoples routes......there's a lot more they haven't all set up the same private server to route too and it only began post server move.....

Your barking up the wrong tree nomad.

You dont know that, have you seen the trace from others having problems being routed thru Europe.?
BTW im not arguing about the data centre move, im not being routed thru Europe (Brisbane here) and my ping has suffered to MWO while it hasnt affected my ping to half a dozen other games i and my sons play, so im believing the problem is MWOs move to cloud servers, after all PGI has said their move was due to the service being cheaper, and im a firm believer that you only get what you pay for, pay peanuts you get monkeys.
All im saying is that alot of the problems with people having funny routes to certain places is because of individuals own routing settings its not an ISP problem but personal settings problems.

[N0MAD]

Kamikaze Viking, on 03 September 2014 - 01:41 AM, said:

Ok then. Here's my trace. From Adelaide on Internode (although it seems they've renamed the infrastructure since the iinet merger)

You can see it goes from Adelaide to Sydney, then California then Canada.

I'm guessing its a Telstra Problem (as always)

C:\Windows\system32>Tracert mwomercs.com

Tracing route to mwomercs.com [192.99.109.192]
over a maximum of 30 hops:

1 2 ms 1 ms 1 ms 192.168.1.1
2 20 ms 20 ms 20 ms lns21.adl2.on.ii.net [203.16.215.199]
3 20 ms 20 ms 22 ms xe10-0-4.cr1.adl2.on.ii.net [150.101.33.203]
4 187 ms 186 ms 186 ms ae0.cr1.adl6.on.ii.net [150.101.33.3]
5 40 ms 40 ms 46 ms ae4.br1.syd7.on.ii.net [150.101.33.34]
6 187 ms 186 ms 186 ms te0-1-1-2.br1.sjc2.on.ii.net [150.101.33.193]
7 219 ms 222 ms 340 ms eqx.snj.ovh.net [206.223.116.193]
8 190 ms 187 ms 189 ms 198.27.73.249
9 * * 240 ms 198.27.73.184
10 * 304 ms 255 ms 198.27.73.179
11 257 ms 257 ms 257 ms bhs-g2-6k.qc.ca [198.27.73.7]
12 253 ms 253 ms 253 ms 198.27.73.233
13 256 ms 256 ms 256 ms mwomercs.com [192.99.109.192]

Trace complete.

My trace from Brisbane is near identical to yours once it hits Sydney with almost identical pings.
Telstra does not set your rout mate, we all goto the Telstra Sydney hub from there the hubs do their work, they will only deviate if they are forced to but its not Telstra setting the routes.
Telstra may eat babies but they dont dictate your internet routes once they hit their international hub, you do that (if you want to)

[N0MAD]

DV McKenna, on 03 September 2014 - 01:26 AM, said:

Your barking up the wrong tree nomad; there are smaller aus ISPs routing the direct way.

Im replying to this since you added to your comment.
So youre telling me you have documented info on how some small ISPs route? can you pls post this info for us?.
I find it hard to believe that ANY Aus ISPs are giving people a EU proxy as part of their setup, but if you have documents that show they are i would apreciate seeing them.

[Oderint dum Metuant]

N0MAD, on 03 September 2014 - 01:46 AM, said:

You dont know that, have you seen the trace from others having problems being routed thru Europe.?
BTW im not arguing about the data centre move, im not being routed thru Europe (Brisbane here) and my ping has suffered to MWO while it hasnt affected my ping to half a dozen other games i and my sons play, so im believing the problem is MWOs move to cloud servers, after all PGI has said their move was due to the service being cheaper, and im a firm believer that you only get what you pay for, pay peanuts you get monkeys.
All im saying is that alot of the problems with people having funny routes to certain places is because of individuals own routing settings its not an ISP problem but personal settings problems.

Have you read the other thread that has been linked and referenced too a dozen times.

I am guessing that because the destination of MWO's server has physically changed albeit the same country; something has gone awry in the route the traffic is taking this is down to either Telstra or their tier 1 partner.

Your other games have not physically moved so their path remains the same as always.

See the postman annology ego used above.

N0MAD, on 03 September 2014 - 02:07 AM, said:

Im replying to this since you added to your comment.
So youre telling me you have documented info on how some small ISPs route? can you pls post this info for us?.
I find it hard to believe that ANY Aus ISPs are giving people a EU proxy as part of their setup, but if you have documents that show they are i would apreciate seeing them.

Use your eyes look in the thread it has traces from Telstra users and iinet and internode(I think that was the name) users towards the end two completely different routes.

Edited by DV McKenna, 03 September 2014 - 02:16 AM.

[applor]

I am in Brisbane on Telstra ADSL2+ and here is my trace route. Previously ping was 240 now sits on 310

Tracing route to www.mwomercs.com [192.99.109.192]
over a maximum of 30 hops:
1 1 ms <1 ms <1 ms 192.168.1.254
2 17 ms 16 ms 17 ms 172.18.212.3
3 16 ms 16 ms 16 ms 172.18.69.101
4 16 ms 16 ms 16 ms bundle-ether4.woo-edge902.brisbane.telstra.net [
203.50.44.42]
5 19 ms 15 ms 17 ms bundle-ether6.woo-core1.brisbane.telstra.net [20
3.50.11.138]
6 31 ms 31 ms 31 ms bundle-ether11.chw-core10.sydney.telstra.net [20
3.50.11.70]
7 32 ms 31 ms 31 ms Bundle-ether17.oxf-gw2.sydney.telstra.net [203.50.13.70]
8 31 ms 31 ms 31 ms bundle-ether1.sydo-core01.sydney.reach.com [203.50.13.38]
9 34 ms 31 ms 31 ms i-0-3-2-0.sydo-core02.bi.telstraglobal.net [202.84.220.189]
10 176 ms 175 ms 175 ms i-0-2-0-2.1wlt-core01.bx.telstraglobal.net [202.84.140.21]
11 173 ms 175 ms 173 ms i-0-4-0-5.eqla01.bi.telstraglobal.net [202.40.149.238]
12 174 ms 171 ms 172 ms tisparkle-peer.eqla01.pr.telstraglobal.net [134.159.61.26]
13 174 ms 173 ms 174 ms xe-9-3-0.paloalto2.pao.seabone.net [89.221.35.141]
14 * * 336 ms pal-5-6k.ca.us [178.32.135.206]
15 * * 291 ms 198.27.73.177
16 306 ms * 307 ms 198.27.73.175
17 319 ms 320 ms 320 ms bhs-g1-6k.qc.ca [198.27.73.1]
18 306 ms 306 ms 306 ms 198.27.73.231
19 305 ms 305 ms 305 ms mwomercs.com [192.99.109.192]

[Oderint dum Metuant]

applor, on 03 September 2014 - 02:10 AM, said:

I am in Brisbane on Telstra ADSL2+ and here is my trace route. Previously ping was 240 now sits on 310

Tracing route to www.mwomercs.com [192.99.109.192]
over a maximum of 30 hops:
1 1 ms &lt;1 ms &lt;1 ms 192.168.1.254
2 17 ms 16 ms 17 ms 172.18.212.3
3 16 ms 16 ms 16 ms 172.18.69.101
4 16 ms 16 ms 16 ms bundle-ether4.woo-edge902.brisbane.telstra.net [
203.50.44.42]
5 19 ms 15 ms 17 ms bundle-ether6.woo-core1.brisbane.telstra.net [20
3.50.11.138]
6 31 ms 31 ms 31 ms bundle-ether11.chw-core10.sydney.telstra.net [20
3.50.11.70]
7 32 ms 31 ms 31 ms Bundle-ether17.oxf-gw2.sydney.telstra.net [203.50.13.70]
8 31 ms 31 ms 31 ms bundle-ether1.sydo-core01.sydney.reach.com [203.50.13.38]
9 34 ms 31 ms 31 ms i-0-3-2-0.sydo-core02.bi.telstraglobal.net [202.84.220.189]
10 176 ms 175 ms 175 ms i-0-2-0-2.1wlt-core01.bx.telstraglobal.net [202.84.140.21]
11 173 ms 175 ms 173 ms i-0-4-0-5.eqla01.bi.telstraglobal.net [202.40.149.238]
12 174 ms 171 ms 172 ms tisparkle-peer.eqla01.pr.telstraglobal.net [134.159.61.26]
13 174 ms 173 ms 174 ms xe-9-3-0.paloalto2.pao.seabone.net [89.221.35.141]
14 * * 336 ms pal-5-6k.ca.us [178.32.135.206]
15 * * 291 ms 198.27.73.177
16 306 ms * 307 ms 198.27.73.175
17 319 ms 320 ms 320 ms bhs-g1-6k.qc.ca [198.27.73.1]
18 306 ms 306 ms 306 ms 198.27.73.231
19 305 ms 305 ms 305 ms mwomercs.com [192.99.109.192]

Hop 14 is france; have you ever modified your connection to make use of a "vps"

I'm guessing not but ya know..