tr1p schreef:

Ik zal dat ook even proberen bij online. Het is vreemd dat dat de oplossing is aangezien het een routerings probleem lijkt.

Ik laat wel weten of het gelukt is.

Daarnaast zal ik aangeven dat er meer mensen zijn met dezelfde problemen.

Dear sir, madam,

Since 3 days some customers from a certain ISP (Online.nl) from the Netherlands, having problems with their ping times to some of your servers located in the UK.

Mostly its the respons times to eu west servers that you host for RIOT Games. But also i find difficulties to achieve normal respons times to servers located in Ireland.

I would like to ask you for some information about this because the first line tech workers of my ISP (online.nl) dont understand a thing on how the internet works.

Its a specific routing problem because customers from other ISP's in the Netherlands dont suffer from this.

Extra info: Normal ping times to your server hosted in the UK are arround 25-35 but since 3 days its anywhere between 250-850 which makes it impossible to play games. For now i got 7 people approaching me with the same problem all got Online.nl as ISP.

Its seems to affect routes from NL > Frankfurt > UK seeing the traceroutes.


I hope you can help us!

R. Trip
Netherlands

Hello Robin,

Can you give us a couple of source IPs back in the Netherlands that I can run MTRs back to from our Frankfurt location please? It's possible it may be a return path issue but I'd like to see.

Also, do you have any copies of some of traces that we may look at as well?

Thank you,

-----------------------------------------------------------------------
Dennis Carrasquillo CCNA NOC

Glad you mentioned Frankfurt. It seem to be the location that gives this problem. If we (users with same problem) use a little program called wtfast than the problem is solved. So in my opinion your suggestion may be right. I think the best IP that i can give you is that of myself and another customer with the same problems.

My ip: 212.64.110.98
Another ip with the same problems is: 83.117.35.88

Both ip's are from online.nl ISP.

Since the IP ranges of RIOT games eu west are not pingable or traceable for me (they ll time out when i reach the important hops) i ll send you a trace route to a facebook.com server. I believe its gets routed in the same way seeing exactly the same pastern in traces and response times. On average the rsp normally is 25-35 now you can see its around 250-700, which gives exactly the same problem.
Here is the ping :
C:\Users\tr1p>ping www.facebook.com /t

Pinging star.c10r.facebook.com [31.13.64.17] with 32 bytes of data:
Reply from 31.13.64.17: bytes=32 time=392ms TTL=88
Reply from 31.13.64.17: bytes=32 time=375ms TTL=88
Reply from 31.13.64.17: bytes=32 time=386ms TTL=88
Reply from 31.13.64.17: bytes=32 time=263ms TTL=88
Reply from 31.13.64.17: bytes=32 time=279ms TTL=88
Reply from 31.13.64.17: bytes=32 time=682ms TTL=88
Reply from 31.13.64.17: bytes=32 time=597ms TTL=88
Reply from 31.13.64.17: bytes=32 time=479ms TTL=88
Reply from 31.13.64.17: bytes=32 time=425ms TTL=88
Reply from 31.13.64.17: bytes=32 time=452ms TTL=88
Reply from 31.13.64.17: bytes=32 time=541ms TTL=88
Reply from 31.13.64.17: bytes=32 time=509ms TTL=88
Reply from 31.13.64.17: bytes=32 time=530ms TTL=88
Reply from 31.13.64.17: bytes=32 time=595ms TTL=88
Reply from 31.13.64.17: bytes=32 time=657ms TTL=88

Here the trace:
C:\Users\tr1p>tracert www.facebook.com

Tracing route to star.c10r.facebook.com [31.13.64.17]
over a maximum of 30 hops:

1 <1 ms <1 ms <1 ms 192.168.1.1
2 98 ms 237 ms 20 ms sd4406e01.adsl.online.nl [212.64.110.1]
3 38 ms 151 ms 20 ms V78.har1e-Asd-O.nl.euro.net [194.134.18.133]
4 18 ms 17 ms 18 ms 194.134.18.134
5 18 ms 17 ms 18 ms ae3.cr1-asd8.nl.euro.net [194.134.161.215]
6 18 ms 17 ms 18 ms ae0.br1-asd8.nl.euro.net [194.134.161.171]
7 18 ms 58 ms 26 ms br01.ams1.tfbnw.net [195.69.145.115]
8 19 ms 18 ms 18 ms ae1.bb01.ams2.tfbnw.net [31.13.29.86]
9 18 ms 18 ms 18 ms ae1.pr02.ams2.tfbnw.net [31.13.29.89]
10 732 ms 717 ms * po126.msw01.02.ams2.tfbnw.net [31.13.24.59]
11 549 ms 420 ms 428 ms edge-star-shv-02-ams2.facebook.com [31.13.64.17]


Trace complete.

I dont know if this is any helpful to you, otherwise you can maybe provide me with some Frankfurt IP addresses that i can trace to see where it goes wrong.

With kind regards,

R. Trip

bij online kom je niet veel verder aangezien de eerstelijns daar er niet veel van begrijpen.

voslex schreef:


Daar ben ik idd ook een beetje bang voor, maar als iemand zegt dat hij na een reset nergens meer last van had...

tr1p schreef:



Die heb ik net gesproken, heeft nog steeds problemen, je doelde toch op Mvrx?

tr1p schreef:



Die heb ik net gesproken, heeft nog steeds problemen, je doelde toch op Mvrx?

Hello Robin,

Sorry for the delay in the response. It's one of those days.

So, I ran some traces back to those IPs. Internap has multiple providers to route out of Frankfurt and couple of the paths shows high latency on the last 2 hops returning. It's actually sort of strange as the same nodes on the paths without that latency are the same as the nodes with the high latency.

For 212.64.110.98 the current outbound path is Verio/NTT. Latency looks fine but I believe that your ISP is also rate-limiting/blocking ICMP as it pinged a couple times and then quit responding.

Host Loss% Snt Last Avg Best Wrst StDev
1. border1.ge3-1-servlan1 0.0% 225 0.3 3.7 0.3 167.1 17.4
2. core2-0-0-0-0.fra002.p 0.0% 225 0.9 0.9 0.6 2.3 0.2
3. 212.119.15.5 0.0% 224 1.0 1.1 0.8 15.8 1.0
4. ae-3.r20.frnkge04.de.b 0.0% 224 0.8 5.2 0.7 71.6 12.6
5. ae-1.r23.amstnl02.nl.b 0.0% 224 7.9 8.6 7.1 16.8 2.3
6. ae-2.r02.amstnl02.nl.b 0.0% 224 8.7 9.0 7.5 15.5 2.3
7. nikhef.openpeering.nl 0.0% 224 20.9 21.3 13.4 24.8 2.5
8. openpeering-gs.online. 0.0% 224 7.8 9.0 7.4 24.3 2.8
9. ae5.cr1-asd8.nl.euro.n 0.0% 224 11.6 9.7 7.8 24.9 2.8
10. PC1.dr6-asd8.nl.euro.n 0.0% 224 14.6 9.8 8.3 22.8 2.5
11. V76.har1e-Asd-Btv.nl.e 0.0% 224 17.8 22.7 17.1 36.9 4.0
12. V77.har1-Asd-Z.nl.euro 0.0% 224 11.9 17.5 11.2 44.8 5.8
13. sd4406e62.adsl.online. 98.7% 224 25.1 25.3 24.9 25.8 0.5

SAME TRACE WITH JUST IPs:
Host Loss% Snt Last Avg Best Wrst StDev
1. 95.172.66.28 0.0% 225 0.3 3.7 0.3 167.1 17.4
2. 95.172.67.4 0.0% 225 0.9 0.9 0.6 2.3 0.2
3. 212.119.15.5 0.0% 224 1.0 1.1 0.8 15.8 1.0
4. 129.250.3.93 0.0% 224 0.8 5.2 0.7 71.6 12.6
5. 129.250.3.179 0.0% 224 7.9 8.6 7.1 16.8 2.3
6. 129.250.2.159 0.0% 224 8.7 9.0 7.5 15.5 2.3
7. 195.69.144.189 0.0% 224 20.9 21.3 13.4 24.8 2.5
8. 82.150.157.75 0.0% 224 7.8 9.0 7.4 24.3 2.8
9. 194.134.161.170 0.0% 224 11.6 9.7 7.8 24.9 2.8
10. 194.134.161.214 0.0% 224 14.6 9.8 8.3 22.8 2.5
11. 194.134.18.129 0.0% 224 17.8 22.7 17.1 36.9 4.0
12. 194.134.18.131 0.0% 224 11.9 17.5 11.2 44.8 5.8
13. 212.64.110.98 98.7% 224 25.1 25.3 24.9 25.8 0.5

For 83.117.35.88 I see the same thing. Some paths are fine on latency but a couple of the other paths are high with latency. The current path from our network to 83.117.35.88 is over GTT/Tinet.

Host Loss% Snt Last Avg Best Wrst StDev
1. border1.ge3-1-servlan1 0.2% 624 0.6 4.4 0.3 220.1 22.3
2. core3-0-0-0-0.fra002.p 0.0% 624 0.9 1.0 0.6 60.3 2.5
3. xe-1-3-1.fra23.ip4.tin 0.0% 624 0.4 2.9 0.3 66.8 9.4
4. te0-4-0-12.mag21.fra03 0.0% 624 1.2 1.1 0.7 59.8 3.2
5. be2044.ccr21.fra03.atl 0.0% 624 1.2 2.6 0.8 508.1 22.6
be2009.mpd22.fra03.atlas.cogentco.com
be2027.ccr22.fra03.atlas.cogentco.com
be2028.mpd21.fra03.atlas.cogentco.com
6. be2263.mpd21.ams03.atl 0.0% 624 8.1 9.3 7.5 242.9 14.4
be2264.mpd22.ams03.atlas.cogentco.com
be2261.ccr21.ams03.atlas.cogentco.com
be2262.ccr22.ams03.atlas.cogentco.com
7. te8-2.ccr01.ams05.atla 0.0% 624 7.9 25.6 7.4 268.4 46.3
te4-1.ccr01.ams05.atlas.cogentco.com
te4-2.ccr01.ams05.atlas.cogentco.com
te3-1.ccr01.ams05.atlas.cogentco.com
te7-2.ccr01.ams05.atlas.cogentco.com
te8-1.ccr01.ams05.atlas.cogentco.com
8. 149.11.80.58 0.0% 623 7.6 8.3 7.5 164.5 7.7
9. ae5.cr1-asd6.nl.euro.n 0.0% 623 7.8 8.5 7.6 66.1 4.6
10. ???
11. V3179.har1e-Ehv-Vkh.nl 0.0% 623 14.9 29.2 13.9 184.7 35.0
12. V3178.har1-Ehv-Woe.nl. 0.8% 623 139.2 208.8 14.0 757.1 180.8
13. s53752358.adsl.online. 88.4% 623 118.9 170.6 31.7 750.5 133.8

Host Loss% Snt Last Avg Best Wrst StDev
1. 95.172.66.28 0.2% 624 0.6 4.4 0.3 220.1 22.3
2. 95.172.67.5 0.0% 624 0.9 1.0 0.6 60.3 2.5
3. 77.67.73.165 0.0% 624 0.4 2.9 0.3 66.8 9.4
4. 130.117.14.85 0.0% 624 1.2 1.1 0.7 59.8 3.2
5. 154.54.74.133 0.0% 624 1.2 2.6 0.8 508.1 22.6
154.54.74.145
154.54.74.141
154.54.74.137
6. 154.54.61.5 0.0% 624 8.1 9.3 7.5 242.9 14.4
154.54.6.209
154.54.37.29
154.54.37.33
7. 154.54.72.53 0.0% 624 7.9 25.6 7.4 268.4 46.3
130.117.0.82
130.117.1.106
130.117.51.33
154.54.72.45
154.54.72.49
8. 149.11.80.58 0.0% 623 7.6 8.3 7.5 164.5 7.7
9. 194.134.161.216 0.0% 623 7.8 8.5 7.6 66.1 4.6
10. ???
11. 194.134.20.109 0.0% 623 14.9 29.2 13.9 184.7 35.0
12. 194.134.20.106 0.8% 623 139.2 208.8 14.0 757.1 180.8
13. 83.117.35.88 88.4% 623 118.9 170.6 31.7 750.5 133.8

This along with the Facebook trace leads me to believe that something is going on at your ISP level on the return path. Mainly because the Faecbook trace doesn't come through Internap at all so you have 2 independent networks showing the same behavior on the return.

I'm including a text file with all the traces I did over all of our providers. Maybe it will help your ISP narrow this down some hopefully.

Thanks for your help anyway! As its not even your job to answer a random email from a non customer. I called in your help because my ISP wont let me trough with the information i gave them. I believe their tech workers are not educated enough to understand what is happening as they are just following the phone scripts they were told to follow.

Somehow i need to get in contact with their noc to narrow this down. As i agree on your thought that its on isp level since non of my friends have this problem using all different ISP's. Add to that that i found several online.nl customers with the exact same problem its seems a waterproof story so far. I will contact them with this information and include your results in it.

Thank you very much for you help finally someone who does understand

With Kinds Regards,

R. Trip