cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 

Routing issue from Calgary/Vancouver to Seattle via Level3

liliass
Helpful Neighbour

Issue : https://www.reddit.com/r/telus/comments/ti9yko/higher_ping/

 

Target IP : 4.14.240.209 (Level3 Seattle) 

 

Problem : High Latency. Up to 60ms.

Expectation : Below 20ms.  

 

Effected users : all Telus gaming users.

I will post treceroute if needed.  

 

 

1 ACCEPTED SOLUTION

Nighthawk
Community Power User
Community Power User

Whether the IP is business or residential, it won't make a difference. They both will use the same path to the server. Keep in mind, all of the routers / devices in question are outside of Telus' network. Level3 is notoriously difficult to deal with if there are problems. 

 

Seattle and Santa Clara/LA are two very different regions. Telus has a fibre backbone that runs to Seattle. After that, Level3 or another backbone provider takes over. Most servers in the LA region have always had higher latency just from distance alone and no ISP can fix that. It's more a matter of physics. Every router or device the connection goes through will add latency. The amount of traffic going through any one of those devices could easily add more unexpected latency too if they get heavily utilized. The Level3 node in Seattle that keeps coming up is likely heavily used so latency to that device will likely be more unpredictable depending on utilization. I always seem to average 50-60ms to that first Level3 device in Seattle from Alberta.

 

Telus most likely has a peering agreement with Level3, but as users we have no idea what the details are. Tech support has zero capability to deal with peering. The front end network support they have access to won't either. There is a very specialized team that deals with peering issues. I've dealt with them before and gotten issues corrected, but you need a huge amount of data for them to work with to try convince Level3 they have a problem. What I've seen so far isn't sufficient for them to look at it. More data is required.

 

Also if you want to post a formatted traceroute like I did, look at the top of the text box for the icons. Click the ellipsis icon:

mce1.png

Then click the code icon:

mce2.png

and paste the traceroute in the box that pops up.

🙂


If you find a post useful, please give the author a "Like" or mark as an accepted solution if it solves your trouble. 🙂

View solution in original post

7 REPLIES 7

liliass
Helpful Neighbour

Target IP : 168.143.191.134 (NTT Seattle). > 60ms as well. 

Nighthawk
Community Power User
Community Power User

Please post the name of the game, the platform it's being played on, and the actual game server IP, not just an intermediate IP. Both IPs you posted are edge routers to their respective company's networks.

 

If there is a latency issue, especially outside of Telus' network, the only way to get any of the companies along the path to the game server in question to even look at the latency is if there is much more data available including the destination IP. This is mainly due to the issue could be almost anywhere along the path and not just what shows on traceroute.

 

Thankfully one user in the reddit post you linked actually including the Genshin server IP in his traceroute screenshots. I have an average of 57ms from Alberta to the L3's Seattle edge router and only 74ms average to the Genshin server. The reddit user who appears to be in the lower mainland has the same latency to the game server but a far lower latency to Seattle in the fixed traceroute. The main latency issue is not that first device. Since the Genshin server is on the US east coast, latency will be higher just due to the distance and to one of the intermediate hops. I'm just curious if that user's latency to Seattle is still that low.

 

Example from Alberta to the Genshin server:

|------------------------------------------------------------------------------------------|
|                                      WinMTR statistics                                   |
|                       Host              -   %  | Sent | Recv | Best | Avrg | Wrst | Last |
|------------------------------------------------|------|------|------|------|------|------|
|                           192.168.1.254 -    0 |  100 |  100 |    0 |    1 |   19 |    0 |
|                             10.137.48.1 -    0 |  100 |  100 |    1 |    3 |   77 |    2 |
|                           154.11.15.111 -    0 |  100 |  100 |   25 |   27 |   81 |   34 |
|          ae57.edge6.Seattle1.Level3.net -    0 |  100 |  100 |   55 |   59 |  140 |   55 |
|                   No response from host -  100 |   21 |    0 |    0 |    0 |    0 |    0 |
|                              4.34.97.90 -   87 |   23 |    3 |    0 |   74 |   76 |   73 |
|                   No response from host -  100 |   21 |    0 |    0 |    0 |    0 |    0 |
|                   No response from host -  100 |   21 |    0 |    0 |    0 |    0 |    0 |
|                   No response from host -  100 |   21 |    0 |    0 |    0 |    0 |    0 |
|                           47.252.29.148 -    0 |  100 |  100 |   73 |   74 |  103 |   73 |
|________________________________________________|______|______|______|______|______|______|

 


If you find a post useful, please give the author a "Like" or mark as an accepted solution if it solves your trouble. 🙂

liliass
Helpful Neighbour

Hi Nighthawk, 

 

Current IP :   23.17.112.X

Game : PUBG MOBILE 

Device : iOS / Android 

Target IP / Host :    49.51.38.210   Tencent San Jose

 

 

traceroute to 49.51.38.210 (49.51.38.210) , 5 relative hops max, 52 byte packets
1 dlinkrouter.(null) (192.168.0.1) 4.099 ms 3.030 ms 2.462 ms
2 10.252.10.1 (10.252.10.1) 4.778 ms 30.389 ms 31.567 ms
3 154.11.15.111 (154.11.15.111) 21.410 ms 23.280 ms 24.197 ms
4 ae57.edge6.seattle1.level3.net (4.14.240.209) 59.859 ms 60.448 ms 59.549 ms
5 * * *
6 4.53.20.6 (4.53.20.6) 70.317 ms 72.155 ms 73.241 ms
7 * * *
8 * * *
9 * * *
10 * * *
11 * * *
12 49.51.38.210 (49.51.38.210) 70.111ms 76.121.ms 79.211ms

 

 

Addition note: I realized some IP range from Telus can go as low as 40ms to the target IP. 
I will post that IP range once I figure it out later. Majority of the IP > 60ms to Level3 Seattle. 

 

liliass
Helpful Neighbour

Ok, Here is another proof, Different IP pool from Telus give you different result of your latency. Here I finally connected to this  142.59.113.X  after several obtain. 

This IP has the lowest latency to Seattle (Level3 / NTT ). But this IP is extremely hard to get. 

 


traceroute to 49.51.38.210 (49.51.38.210) , 5 relative hops max, 52 byte packets
1 dlinkrouter.(null) (192.168.0.1) 12.172 ms 15.880 ms 18.671 ms
2 10.252.10.1 (10.252.10.1) 6.355 ms 7.694 ms 8.811 ms
3 154.11.15.109 (154.11.15.109) 22.829 ms 25.440 ms 27.053 ms
4 ae57.edge6.seattle1.level3.net (4.14.240.209) 25.294 ms 28.216 ms 30.344 ms
5 * * *
6 4.53.20.6 (4.53.20.6) 49.615 ms 54.262 ms 57.850 ms
7 * * *
8 * * *
9 * * *
10 * * *
11 * * *
12 49.51.38.210 (49.51.38.210) 43.542ms 45.131.ms 48.911ms

 

 

Below is another traceroute to GcoreLabs IP located in Santa Clara via NTT Seattle. 

traceroute to 92.38.148.2 (92.38.148.2) , 5 relative hops max, 52 byte packets
1 dlinkrouter.(null) (192.168.0.1) 3.730 ms 7.029 ms 9.733 ms
2 10.252.10.1 (10.252.10.1) 5.195 ms 6.763 ms 8.000 ms
3 154.11.2.81 (154.11.2.81) 23.299 ms 25.477 ms 26.473 ms
4 ae-9.a03.sttlwa01.us.bb.gin.ntt.net (168.143.191.134) 24.779 ms 27.106 ms 28.342 ms
5 ae-3.r25.sttlwa01.us.bb.gin.ntt.net (129.250.2.206) 26.728 ms 28.945 ms 31.204 ms
6 ae-3.r25.snjsca04.us.bb.gin.ntt.net (129.250.3.124) 41.233 ms 44.269 ms 46.113 ms
7 ae-41.r02.snjsca04.us.bb.gin.ntt.net (129.250.6.119) 41.429 ms 44.265 ms 45.983 ms
8 * * *
9 * * *
10 * * *
11 gw.gcore.lu (5.8.24.65) 49.797 ms 85.557 ms 90.478 ms
12 vrrp.gcore.lu (92.38.148.2) 43.312 ms 55.255 ms 60.078 ms

liliass
Helpful Neighbour

Please let me emphasize this,  only 142.59.113.X IP can get decent low latency to Seattle/Santa Clara/LA or any city located in US West. Unless route via SeattleIX, PCCW Seattle both 20ms for all IP’s 

Im not sure why is this happening, maybe QoS / Priority level / This IP belongs to business user. 

 

But majority of the IP address give you extra 30-40ms to US West Coast.

Hopefully Telus can fix this asap. Thank you 

Nighthawk
Community Power User
Community Power User

Whether the IP is business or residential, it won't make a difference. They both will use the same path to the server. Keep in mind, all of the routers / devices in question are outside of Telus' network. Level3 is notoriously difficult to deal with if there are problems. 

 

Seattle and Santa Clara/LA are two very different regions. Telus has a fibre backbone that runs to Seattle. After that, Level3 or another backbone provider takes over. Most servers in the LA region have always had higher latency just from distance alone and no ISP can fix that. It's more a matter of physics. Every router or device the connection goes through will add latency. The amount of traffic going through any one of those devices could easily add more unexpected latency too if they get heavily utilized. The Level3 node in Seattle that keeps coming up is likely heavily used so latency to that device will likely be more unpredictable depending on utilization. I always seem to average 50-60ms to that first Level3 device in Seattle from Alberta.

 

Telus most likely has a peering agreement with Level3, but as users we have no idea what the details are. Tech support has zero capability to deal with peering. The front end network support they have access to won't either. There is a very specialized team that deals with peering issues. I've dealt with them before and gotten issues corrected, but you need a huge amount of data for them to work with to try convince Level3 they have a problem. What I've seen so far isn't sufficient for them to look at it. More data is required.

 

Also if you want to post a formatted traceroute like I did, look at the top of the text box for the icons. Click the ellipsis icon:

mce1.png

Then click the code icon:

mce2.png

and paste the traceroute in the box that pops up.

🙂


If you find a post useful, please give the author a "Like" or mark as an accepted solution if it solves your trouble. 🙂

liliass
Helpful Neighbour

Hi Nighthawk, 

Thanks your rapid reply. Regarding of your answer

  • Level3 is notoriously difficult to deal with if there are problems.
  • Perhaps, The NOC team / Peering team can change to a better node/route. (Eg. via SeattleIX, Telus has 100G peering with them.) 

 

  • Seattle and Santa Clara/LA are two very different regions. 
  • I do understand, but all US West Coast route must route via Seattle Level3/NTT. As a result, We (as a gamer) are having extra 30-40ms compare to players with other ISP. Telus rated as Best ISP for Gaming in Canada. 

 

  • Why 142.59.113.X has a lower latency compare to other IP Pool ? Perhaps, The NOC/Peering/Network team can figure it out. And try to fix it ? 

Lastly, I think this issue must be escalate to NOC or Peering team for future investigation. 

 

Thank you.