Busy testing the new upstream paths across a wider range of prefixes today, seeing the latency and performance issue. Just running a few tests before rolling back. Apologies for the inconvenience.
-
Join the MyBroadband community
South Africa’s biggest forum. Discuss, discover, and connect with thousands of members.
-
Question of the Day: What's the most interesting small town you've visited in South Africa?
Web Squad ISP
- Thread starter Gazed416
- Start date
- Status
Busy testing the new upstream paths across a wider range of prefixes today, seeing the latency and performance issue. Just running a few tests before rolling back. Apologies for the inconvenience.
Ahh, that explains it, I'll close the ticket then!
Gaming feels quite responsive. If it wasn't for the packet loss it would be pretty nice.
Ton of packet loss still @websquadza - is the testing ongoing?
Code:
|------------------------------------------------------------------------------------------|
| WinMTR statistics |
| Host - % | Sent | Recv | Best | Avrg | Wrst | Last |
|------------------------------------------------|------|------|------|------|------|------|
| 192.168.1.1 - 0 | 26 | 26 | 0 | 0 | 0 | 0 |
| core.as-01.cp1.za.ws.net.za - 0 | 26 | 26 | 0 | 0 | 1 | 0 |
| No response from host - 100 | 5 | 0 | 0 | 0 | 0 | 0 |
| core.pe-xe-ip01.cp1.za.ws.net.za - 0 | 26 | 26 | 0 | 2 | 52 | 1 |
| 165-69-148-197.as37497.za.net - 0 | 26 | 26 | 0 | 2 | 51 | 1 |
| 162-68-148-197.as37497.za.net - 0 | 26 | 26 | 0 | 2 | 51 | 1 |
| 14-71-148-197.as37497.za.net - 12 | 17 | 15 | 143 | 143 | 144 | 143 |
| 197.148.66.41 - 12 | 17 | 15 | 143 | 143 | 144 | 143 |
|ge-0-1-0-68.peering1.the.lon1.coreix.net - 12 | 17 | 15 | 144 | 145 | 159 | 144 |
| e0-core4.enf.coreix.net - 12 | 17 | 15 | 144 | 144 | 145 | 144 |
| m12-ai04.enf.coreix.net - 12 | 17 | 15 | 144 | 144 | 146 | 144 |
| No response from host - 100 | 5 | 0 | 0 | 0 | 0 | 0 |
| 85.13.221.178.reverse.coreix.net - 12 | 17 | 15 | 144 | 144 | 145 | 144 |
|________________________________________________|______|______|______|______|______|______|
WinMTR v0.92 GPL V2 by Appnor MSP - Fully Managed Hosting & Cloud ProviderAll along the route:
Code:
|------------------------------------------------------------------------------------------|
| WinMTR statistics |
| Host - % | Sent | Recv | Best | Avrg | Wrst | Last |
|------------------------------------------------|------|------|------|------|------|------|
| 192.168.1.1 - 0 | 48 | 48 | 0 | 0 | 0 | 0 |
| core.as-01.cp1.za.ws.net.za - 0 | 48 | 48 | 0 | 0 | 1 | 0 |
| No response from host - 100 | 10 | 0 | 0 | 0 | 0 | 0 |
| core.pe-xe-ip01.cp1.za.ws.net.za - 3 | 44 | 43 | 0 | 1 | 6 | 1 |
| 165-69-148-197.as37497.za.net - 3 | 44 | 43 | 0 | 0 | 1 | 0 |
| 162-68-148-197.as37497.za.net - 3 | 44 | 43 | 0 | 0 | 2 | 0 |
| 14-71-148-197.as37497.za.net - 3 | 44 | 43 | 143 | 143 | 144 | 143 |
| 197.148.66.41 - 3 | 44 | 43 | 143 | 143 | 144 | 143 |
| be20.asr01.ld5.as20860.net - 5 | 40 | 38 | 145 | 145 | 146 | 145 |
| be10.asr02.dc5.as20860.net - 5 | 40 | 38 | 145 | 145 | 146 | 145 |
| 130.180.203.108 - 5 | 40 | 38 | 145 | 145 | 146 | 145 |
| 1717.g1.1ug.dc5.as20860.net - 9 | 36 | 33 | 145 | 149 | 216 | 150 |
| 217.147.89.101 - 3 | 44 | 43 | 144 | 145 | 183 | 144 |
|________________________________________________|______|______|______|______|______|______|
WinMTR v0.92 GPL V2 by Appnor MSP - Fully Managed Hosting & Cloud ProviderWe’ve rolled back changes, but something is still out. We’re investigating and will revert ASAP.
CPT general notice. We’re going to bounce a few bgp sessions to see if the issue is not due to stale connections. Services will be briefly interrupted. We apologise for the inconvenience. Please expect a few drops over the next few minutes.
Sessions bounced. Monitoring closely
Seem to be getting way higher latency than usual to some Apex server locations in EU. Could just be Apex though. Here's an example:
Code:
|------------------------------------------------------------------------------------------|
| WinMTR statistics |
| Host - % | Sent | Recv | Best | Avrg | Wrst | Last |
|------------------------------------------------|------|------|------|------|------|------|
| 192.168.1.1 - 0 | 11 | 11 | 0 | 0 | 0 | 0 |
| core.as-01.cp1.za.ws.net.za - 0 | 11 | 11 | 0 | 0 | 1 | 1 |
| No response from host - 100 | 2 | 0 | 0 | 0 | 0 | 0 |
| core.pe-xe-ip01.cp1.za.ws.net.za - 0 | 11 | 11 | 0 | 1 | 2 | 1 |
| 165-69-148-197.as37497.za.net - 0 | 11 | 11 | 0 | 0 | 1 | 1 |
| 162-68-148-197.as37497.za.net - 0 | 11 | 11 | 0 | 0 | 2 | 1 |
| 14-71-148-197.as37497.za.net - 0 | 11 | 11 | 143 | 144 | 150 | 143 |
| 197.148.66.41 - 0 | 11 | 11 | 143 | 144 | 149 | 144 |
| linx-224.inx.as13213.net - 0 | 11 | 11 | 144 | 146 | 165 | 144 |
| no-ptr.midphase.com - 0 | 11 | 11 | 220 | 220 | 221 | 220 |
|et-0-1-0.bdr2-a21-lon1-uk.hosting.thg.com - 0 | 11 | 11 | 220 | 222 | 235 | 220 |
|et-0-1-1.bdr1-6525-ams1-nl.hosting.thg.com - 0 | 11 | 11 | 220 | 220 | 230 | 220 |
| 46.23.79.129 - 0 | 11 | 11 | 219 | 219 | 220 | 220 |
| 109.123.102.62 - 0 | 11 | 11 | 219 | 220 | 224 | 219 |
|________________________________________________|______|______|______|______|______|______|
WinMTR v0.92 GPL V2 by Appnor MSP - Fully Managed Hosting & Cloud Provider
Last edited:
Still experiencing this.
@websquadza - 90% of servers I'm joining on Apex today are > 200ms ping, in EU. Pretty much all AWS is 220ms or above.
Weirdly enough, this isn’t to and AWS destination. Everything seems clear to London, then it seems servers are being advertised by some network called UK2 limited, where almost 80ms is added. Are you sure about this server IP above?
PS. The new transit isn’t currently in the mix for production FTTH prefixes and we aren’t accepting routes from them (I’m waiting for some feedback on the 10-15ms higher latencies we’re picking up from them)- so technically, we’re on the same routing we were yesterday morning.
Tested a bunch of streams on our CPT vm, not struggling to load 1080p60.
Something weird's going on on my side then. I couldn't even stream in 720p on Twitch. When I open Apex, I'm suddenly seeing Singapore as the best latency server in the list where before it wouldn't even feature, it was always Bahrain (doesn't come up anymore) then London, Amsterdam, etc. If I connect to it (Singapore) and play a game, >230ms. If I switch to London, 1 server in 10 will be normal 140-170ms, the rest will be >220ms.
Let me try capture all of the TCP connections Apex is making when I get on a high latency server and see if I can isolate which is the game server itself.
Let me try capture all of the TCP connections Apex is making when I get on a high latency server and see if I can isolate which is the game server itself.
Sounds like apex...I get the same on my Afrihost connection. All other games its fine
Sorry for the late reply here. This is an odd one. We don’t have different routing for different prefixes within a region. Bumping your connection would give you another IP within the same prefix pool- which means your routing remains identical. Definitely something with the matchmaking algo.
Weird thing is these secondary servers seem to hop between two routing paths on their way to LINX as well. Doing a little more digging to see what (if anything) can be done there.
- Status