South african servers lag

"

RuanBEAST#3065 wrote:

thats not how networks function, networks dont inherently run with packet loss, but im fine accepting that some packet loss wouldnt affect the game to the degree that it is being affected, however the 2000ms of latency is definitely a point of concern.

I am not excusing the existence of packet loss. You are;

"

RuanBEAST#3065 wrote:

my home network dropped 2 of 480 packets
but if you look at point 9 and 10 you can see 4% of packet loss, which is significantly worse than the 2 packets dropped within my home network.

---

If you have some packet loss plaguing an internet connection - and starting at, or prior to, your home router, you do - then you will naturally see the precise packet loss percentage fluctuate marginally from hop-to-hop. It is not unlike how on average you will get each result of a coin flip 50% of the time - but if you flip a coin ten times, you're not guaranteed each result happening five times.

It is simply a function of a meagre sample size and variance.

If you resolved the packet loss - and, indeed, the spike to >100 ms latency - within your home network, you would almost certainly find that the packet loss seen later in the report would vanish. Indeed, if you kept running the test every hour or so, you would see inconsistencies between the various tests regarding which hops have "the worst" packet loss.

---

It doesn't make sense to investigate symptoms of latency spikes in-game with proven packet loss on your home network that you are seemingly unwilling to resolve. The packet loss will be artificially causing latency spikes and so it will be challenging - if not impossible - to know whether there are actual latency spikes or just dropped packets being re-sent. The first step of methodical troubleshooting is always to identify and remove potential compounding factors.

I don't see a reason to care about how routers respond to ICMP echo requests as playing Path of Exile entails neither ICMP echo requests nor sending traffic of any kind to routers. They quite literally utilise different hardware for forwarding traffic (their job) and responding to traffic (not their job). That only matters if you identify an issue affecting the final hop and wish to determine it begins.

This is because the router's CPU processing pings is not likely to be representative of the ASIC forwarding traffic. There's no reason to assume that it would be. Only in the context of ~identical symptoms across multiple consecutive hops might we think that it is.

---

tl;dr: Your home network has packet loss. Your game client occasionally records latency spikes up to four digits as packets get sent; no acknowledgment is received; and so packets are re-sent seconds later. You're asking us to investigate the symptom while, in the same breath, you ask us to ignore the cause. You're also casually informing us that we don't understand how networks operate...

Irony is dead.