South Africa’s biggest forum. Discuss, discover, and connect with thousands of members.
Where you actually situated? Durban?Not sure if its related but I'm with herotel fiber and also getting massive latency in the evening to the local path of exile servers in jhb, couple of guildies are also experiencing problems, mostly from outside of jhb, CT, Durban etc and on different ISP's.
It usualy starts around 20:00 and continues till around 23:00 after which it's fine again.
When I do a WinMTR the high latency starts on a 196.60.*.* ip address which seems to be from NAPAFRICA-JB, the internet exchange. So not sure if its maybe getting overloaded at night or something, but I don't know enough about how that all works to know for sure.
Cape Town.Where you actually situated? Durban?
You'd think someone like NAPAfrica would be on top of this, but its been weeks.We have been seeing the same and have been troubleshooting the issue extensively with some of our big peers like Rain, Cloudflare, HE and Fastly.
We suspect a portion of the JHB NAP Peering fabric is experiencing congestion in the evenings between ~20:00 and 22:30
You'd think someone like NAPAfrica would be on top of this, but its been weeks.
Happy to test across our interfaces and see if we pick it up too. Let me know where you’re checking and which fabric(s) you’re seeing this on.It's a very tricky issue because it's not affecting all peers and not all interfaces. We only see it on the one 200Gbps part of our peering capacity. What is also interesting is that other ISPs are affected but it's quiet on the NOG groups even when I first raised the issue.
That makes sense in regards to my Path of Exile server issue because you would connect to a instance on an ip and it will be laggy but if you leave and rezone and get lucky with a new route then the same instance with the same ip is 100% fine. I'm just glad its a known issue among the ISP's in the hope that something will be done soon. Really frustrating problem.It's a very tricky issue because it's not affecting all peers and not all interfaces. We only see it on the one 200Gbps part of our peering capacity. What is also interesting is that other ISPs are affected but it's quiet on the NOG groups even when I first raised the issue.
It's a very tricky issue because it's not affecting all peers and not all interfaces. We only see it on the one 200Gbps part of our peering capacity. What is also interesting is that other ISPs are affected but it's quiet on the NOG groups even when I first raised the issue.
Any ETA or already done?Teraco has confirmed that one of the Leaf switches is an issue. They are working on a solution.
Any ETA or already done?
@AfriNatic Seems like this problem is back again, lots of packet loss on the AfriNIC exchange causing high latency for people from CT. Are you seeing anything on your side?We don't have confirmation that the issue is resolved yet but latency during evening peaks have been looking perfect.