Packet Loss - WebAfrica

[Fuzz-O]

Hey All

For the past few months my family's Overwatch 2 and WoW latency has been around 180-210ms. I chalked it up to infrastructure/location changes at Blizzard but previously I was at a solid and playable 120-130ms. Last night I was motivated to check what's up and my PingPlotter results look dismal. The packet loss is astounding. Is it worth reporting this to WA support and hoping an agent will have someone attend or is 180-210 going to be my new normal?

 

[CrypticZA]

Hey All

For the past few months my family's Overwatch 2 and WoW latency has been around 180-210ms. I chalked it up to infrastructure/location changes at Blizzard but previously I was at a solid and playable 120-130ms. Last night I was motivated to check what's up and my PingPlotter results look dismal. The packet loss is astounding. Is it worth reporting this to WA support and hoping an agent will have someone attend or is 180-210 going to be my new normal?

View attachment 1750544

120-130 is probably you playing on the Middle East servers - this route has been very iffy and has been going down quite a lot if i recall. It is an Amazon network route so not controlled by the ISP, Amazon controls that route as its their network being used.

The 180-210 would be Europe servers. I assume with that latency you are sitting in JHB/PTA (up country) which is quite a normal ping to the Blizzard servers.

In terms of that trace route the last hop doesn't respond so you can't say if there is packet loss. Your title hints towards packet loss but you comparing ping.

In terms of route you going JHB -> CT -> WACS cable -> London -> Blizzard which is the correct way

 

[CrypticZA]

120-130 is probably you playing on the Middle East servers - this route has been very iffy and has been going down quite a lot if i recall. It is an Amazon network route so not controlled by the ISP, Amazon controls that route as its their network being used.

The 180-210 would be Europe servers. I assume with that latency you are sitting in JHB/PTA (up country) which is quite a normal ping to the Blizzard servers.

In terms of that trace route the last hop doesn't respond so you can't say if there is packet loss. Your title hints towards packet loss but you comparing ping.

In terms of route you going JHB -> CT -> WACS cable -> London -> Blizzard which is the correct way

As reference this is from Cool Idea's JHB looking glass server ping WA ping is within a couple ms of it to that same IP

traceroute to 64.224.26.160 (64.224.26.160), 15 hops max, 60 byte packets
1 sfp-tengig-1-610-vm.studcar1-coolideas.cisp.co.za (154.0.6.1) [AS37680] 0.546 ms
2 bond-stuas1.stucr1.cisp.co.za (100.99.3.241) [*] 0.482 ms
3 bond-terjhbcs0-570-jb3-gallo-1sturdee-la.terjhb1cr1.cisp.co.za (100.99.2.163) [*] 1.002 ms
4 lo0.trijb3br2.cisp.co.za (JB3/BR2) [AS37680] 1.058 ms
5 bond-cr1-a-1640-wiocc-wacs.ci-the-cr1.cisp.co.za (100.98.1.183) [*] 159.297 ms
6 bond-thecs0.thecr0.cisp.co.za (100.98.1.102) [*] 159.874 ms
7 195.66.226.234 (195.66.226.234) [*] 161.063 ms
8 ae1-br01-eqld5.as57976.net (137.221.79.33) [AS57976] 351.501 ms
9 *
10 *
11 et-0-0-1-pe03-eqam3.as57976.net (137.221.78.67) [AS57976] 168.063 ms
12 137.221.66.41 (137.221.66.41) [AS57976] 167.441 ms
13 *
14 *
15 *

 

[Fuzz-O]

120-130 is probably you playing on the Middle East servers - this route has been very iffy and has been going down quite a lot if i recall. It is an Amazon network route so not controlled by the ISP, Amazon controls that route as its their network being used.

The 180-210 would be Europe servers. I assume with that latency you are sitting in JHB/PTA (up country) which is quite a normal ping to the Blizzard servers.

In terms of that trace route the last hop doesn't respond so you can't say if there is packet loss. Your title hints towards packet loss but you comparing ping.

In terms of route you going JHB -> CT -> WACS cable -> London -> Blizzard which is the correct way

Surely that high amount of failed echo requests is a simple indicator that something is not right, especially at only 5 hops and 73% of them lost.

 

[CrypticZA]

Surely that high amount of failed echo requests is a simple indicator that something is not right, especially at only 5 hops and 73% of them lost.

Router that the traffic is going through has high volumes and is set to not prioritise ICMP pings.

If the packets were being dropped the same value would be seen carried through and if you look at the next hop the packet loss % is 0.

To add to this the final end point hasn't responded to 1 packet so technically if that were the case you would have no connection to the server at all

 

[Fuzz-O]

The must be using Cisco Meraki equipment then . Kidding aside, guess it's my new norm then. thanks for the responses.

 

[CrypticZA]

The must be using Cisco Meraki equipment then . Kidding aside, guess it's my new norm then. thanks for the responses.

You can look at something like Exitlag (gaming VPN) which will connect in JHB and have a stable direct route to Middle East. The Amazon one goes down frequently from what i know.

 

[Fuzz-O]

You can look at something like Exitlag (gaming VPN) which will connect in JHB and have a stable direct route to Middle East. The Amazon one goes down frequently from what i know.

My son uses Exit Lag for Rainbow Six Siege and swears by it. Maybe I'll give it a shot.

 

[cavedog]

The biggest issues with these game networks is they focus their network build on reachability in that region eg EU.

This is problematic for South African networks that learn these routes wherever they land for eg in London, Amsterdam, Marseille, Frankfurt etc and then goes onto their carrier to the actual hosted location.

It's very hard to tweak it to get the lowest latency for SA networks.

Riot Games and Blizzard is the worst.