Huge latncy between Softlayer and Zayo in Dallas US
|
I cut out the first few hops in the post below (not relevant).
This has been a persistent problem for me over the last month or two, but it's been much worse in the last two days. This one peering point, in the same city (possibly building) as PoE's servers, is causing over 100ms of latency. Yesterday this peering point was spiking up to 900ms on a regular prolonged basis. My ping is usually 200ms. FYI, networklayer.com is Softlayer (IBM), which is PoE/GGG's hosting provider. So, anyone saying that this isn't GGG's problem is wrong. They are paying Softlayer, and it's Softlayer's problem if they have a congested peering point. I would be surprised if this isn't already on the radar of Softlayer. Zayo isn't well known for congestion avoidance in their network. They are a "budget" transport provider. |------------------------------------------------------------------------------------------| | WinMTR statistics | | Host - % | Sent | Recv | Best | Avrg | Wrst | Last | |------------------------------------------------|------|------|------|------|------|------| | ae10.cr1.sjc2.us.zip.zayo.com - 0 | 34 | 34 | 38 | 41 | 68 | 39 | | ae0.cr1.lax112.us.zip.zayo.com - 3 | 34 | 33 | 45 | 47 | 63 | 48 | | ae2.cr1.iah1.us.zip.zayo.com - 3 | 34 | 33 | 685 | 727 | 776 | 756 | | ae0.cr1.dfw2.us.zip.zayo.com - 3 | 34 | 33 | 87 | 93 | 112 | 90 | | ae11.er1.dfw2.us.zip.zayo.com - 0 | 34 | 34 | 78 | 80 | 88 | 80 | |64.125.192.214.IPYX-102621-930-ZYO.zip.zayo.com - 3 | 34 | 33 | 679 | 723 | 769 | 768 | | ae1.dar02.sr01.dal01.networklayer.com - 3 | 34 | 33 | 186 | 196 | 209 | 199 | | po2.fcr05.sr06.dal01.networklayer.com - 0 | 33 | 33 | 477 | 514 | 547 | 538 | | patch101.pathofexile.com - 0 | 33 | 33 | 216 | 233 | 261 | 256 | |________________________________________________|______|______|______|______|______|______| WinMTR - 0.8. Copyleft @2000-2002 Vasile Laurentiu Stanimir ( stanimir@cr.nivis.com ) It's interesting how the ping from Zayo's routers are high, but not transit latency is not affected. Could be the result of some QoS (DDoS mitigation) or an active DDoS. I've not seen that behavior before in this route. | |
|
It's not GGG's issue.
I know that's not what you want to hear, but it's the answer you're going to get. You need to email your ISP and have them escalate the issue through their channels. The issue is a peering issue, and it's likely a peering issue on the Zayo side. So contact your ISP and provide them a copy of the trace and ask them to forward it to a network technician to escalate. Good luck. |
