Web Squad ISP

[websquadza]

Evening All. Apologies for my delay here. Please remember that there is an official support channel - support@websquad.co.za. This thread is not monitored 24/7 and we do not get notifications when you post here; so we can't react as quickly as the ticket system allows. If you don't get a quick response here, please log a ticket - teams are monitoring the ticketing system at all times.

That said, we are looking into the issue. Traffic is jumping between two peering routers resulting in these drops; will have an update asap. Apologies for the inconvenience.

 

[Looney]

This on a 1 gig line. Wired. Not cool. Not cool at all.

Edit: used to get over 850MB/s download and 200Mb/s upload.

 

[Senshi]

Evening All. Apologies for my delay here. Please remember that there is an official support channel - support@websquad.co.za. This thread is not monitored 24/7 and we do not get notifications when you post here; so we can't react as quickly as the ticket system allows. If you don't get a quick response here, please log a ticket - teams are monitoring the ticketing system at all times.

That said, we are looking into the issue. Traffic is jumping between two peering routers resulting in these drops; will have an update asap. Apologies for the inconvenience.

Any feedback?

 

[websquadza]

This on a 1 gig line. Wired. Not cool. Not cool at all.

Edit: used to get over 850MB/s download and 200Mb/s upload.

Maybe just give everything a bump (reboot) there; shouldn't be struggling to get your 1G. Getting line wire on my tests here. And that device is hard wired as far as I remember?

 

[websquadza]

Any feedback?

You'll see we've stopped the drops. But we're still investigating to see what the cause has been; so I'm a little wary of saying we're good again until we know 100% we are. In the meanwhile, things are looking a lot more stable over the past 45 mins.

 

[Krazee]

This on a 1 gig line. Wired. Not cool. Not cool at all.

Edit: used to get over 850MB/s download and 200Mb/s upload.

That test is in MB/s not Mbits/s so your getting around 600 down and the full 200 up

 

[abudabi]

I don't think anyone has gotten pitchforks yet and your comment seems to want to stir something there isn't.

I could live with reduced speed or increased latency a little, but packet loss to the point where my internet experience becomes unusable is an issue. I've reported it and am patiently waiting on a repsonse. No one here has brought any pitchforks, a lot of members here post from a genuine interest/concern point of view.

I've recommended Websquad to many of my friends and family and will continue to do so, simply because of the transparency around issues and the quick turnarounds.

LOL dude. I'm as chilled as Mary with her little lamb and not stirring anything. I'm just annoyed by people jumping on the "what the hell is going on" and the "you better sort this out" bandwagon when there's an issue. I've seen the same in the CI thread and it grates me.

 

[Senshi]

LOL dude. I'm as chilled as Mary with her little lamb and not stirring anything. I'm just annoyed by people jumping on the "what the hell is going on" and the "you better sort this out" bandwagon when there's an issue. I've seen the same in the CI thread and it grates me.

Fair enough but that's not getting the pitchforks

 

[websquadza]

JHB Notice: We're conducting some maintenance and updates, you may experience intermittent drops over the next 10-15 minutes. This is all in an effort to improve the network.

 

[websquadza]

JHB Notice: We're conducting some maintenance and updates, you may experience intermittent drops over the next 10-15 minutes. This is all in an effort to improve the network.

Maintenance complete. We are monitoring everything very closely.

 

[TwinDucT]

Any known issues on Vumatel Florida?

Ping plotter showing a bunch of red

Target Name: www.google.com
         IP: 172.217.170.68
  Date/Time: 2021/12/12 15:17:01 - 2021/12/12 15:27:01

Hop  Sent  PL%   Min     Max   Avg  Host Name / [IP]
  1   244    0  0,21   20,99  0,82  192.168.0.1 [192.168.0.1]
  2   244    9  1,63   15,53  3,08  core.bng-xe-02.jb1.za.wecom.net.za [160.119.230.1]
  3   244    9  1,88   13,97  3,58  100.99.198.1 [100.99.198.1]
  4   244    9  1,96   83,32  4,28  core.pe-xe-ip01.jb1.za.ws.net.za [160.119.224.19]
  5   244    5  2,43   87,14  5,05  google.ixp.joburg [196.60.8.166]
  6   244   11  2,42   84,80  4,70  74.125.245.193 [74.125.245.193]
  7   244    9  2,87  108,06  6,30  216.239.40.199 [216.239.40.199]
  8   244    8  2,16  111,09  5,21  www.google.com [172.217.170.68]

 

[websquadza]

Any known issues on Vumatel Florida?

Ping plotter showing a bunch of red

Target Name: www.google.com
         IP: 172.217.170.68
  Date/Time: 2021/12/12 15:17:01 - 2021/12/12 15:27:01

Hop  Sent  PL%   Min     Max   Avg  Host Name / [IP]
  1   244    0  0,21   20,99  0,82  192.168.0.1 [192.168.0.1]
  2   244    9  1,63   15,53  3,08  core.bng-xe-02.jb1.za.wecom.net.za [160.119.230.1]
  3   244    9  1,88   13,97  3,58  100.99.198.1 [100.99.198.1]
  4   244    9  1,96   83,32  4,28  core.pe-xe-ip01.jb1.za.ws.net.za [160.119.224.19]
  5   244    5  2,43   87,14  5,05  google.ixp.joburg [196.60.8.166]
  6   244   11  2,42   84,80  4,70  74.125.245.193 [74.125.245.193]
  7   244    9  2,87  108,06  6,30  216.239.40.199 [216.239.40.199]
  8   244    8  2,16  111,09  5,21  www.google.com [172.217.170.68]

This looks last mile/backhaul related (starting on the hop between your router and our BNG). Have you tried bypassing your router to eliminate that? Please log a ticket

 

[semaphore]

This looks last mile/backhaul related (starting on the hop between your router and our BNG). Have you tried bypassing your router to eliminate that? Please log a ticket

Have you heard anything from evotel of outages in north riding ?

 

[TwinDucT]

This looks last mile/backhaul related (starting on the hop between your router and our BNG). Have you tried bypassing your router to eliminate that? Please log a ticket

Trying to log a ticket but yeah can barely get myadsl to load. Not a 100% sure how I can bypass the router. I have plugged in a new router and it's still the same (Also checked against a wireless device, which had my upstream as 0.50. Which is obviously wrong)

Edit: Finally got a ticket logged

 

[Zodiak]

I have an interesting question, why do I have a lower ping to Hong Kong than to Europe or the Middle East?

 

[CrypticZA]

I have an interesting question, why do I have a lower ping to Hong Kong than to Europe or the Middle East?

Because its going via the SAFE cable which gives around 110ms - 140ms (depends what region you in) to Singapore and then from there it goes to Hong Kong, from CT to Hong Kong you get around 170 via that route

 

[Zodiak]

Thanks, it's weird because in Escape from Tarkov, Hong Kong is at around 170ms and if I run a Speedtest to Hong Kong it is almost 300ms

 

[CrypticZA]

Thanks, it's weird because in Escape from Tarkov, Hong Kong is at around 170ms and if I run a Speedtest to Hong Kong it is almost 300ms

Depends on peering, if the quick route is being used etc.

Traffic on Amazon/Google/Microsoft for example all are picked up locally so they will have poor latency since it goes via their network but if a certain route can go direct it will have the lower latency.

Some networks only peer at certain exchanges globally as well so you might find most of the speedtest servers are peering in London vs in their own region since the company is based in location X but has speedtest servers in many other regions.

So for example in Apex legends there are Google servers/Amazon servers and then another host all in Singapore, the AWS/Google servers you would get poor ping to but the other host has a direct route so you get the low latency.

 

[Zodiak]

Makes sense, thank you!

 

[websquadza]

Depends on peering, if the quick route is being used etc.

Traffic on Amazon/Google/Microsoft for example all are picked up locally so they will have poor latency since it goes via their network but if a certain route can go direct it will have the lower latency.

Some networks only peer at certain exchanges globally as well so you might find most of the speedtest servers are peering in London vs in their own region since the company is based in location X but has speedtest servers in many other regions.

So for example in Apex legends there are Google servers/Amazon servers and then another host all in Singapore, the AWS/Google servers you would get poor ping to but the other host has a direct route so you get the low latency.

I have an interesting question, why do I have a lower ping to Hong Kong than to Europe or the Middle East?

@CrypticZA pretty much summed it up perfectly. These differences are less pronounced in peering heavy regions such as the EU, but in Asia, it really does come down to the upstream providers different peers have. SAFE is relatively new (than the established wavelength providers to Europe) and undersubscribed (probably due to cost), which means most Tier 1s and cloud providers still traverse via Europe to reach South Africa. Our capacity runs over a fairly established Tier 1 into Singapore and Hong Kong, but it does not always mean the return path one of these peers/clouds necessarily sees this as the quickest path back.

And this is one of the caveats of BGP. It's an incredible protocol and the entire internet runs on it - but it does not understand, nor measure latency or path stability (packet loss) for that matter. It starts off by measuring AS hops. Eg. Google peers with Web sQuad in JHB (1 AS hop = shorter path) - this is how Google sees our network globally, 1 hop. So, a Google Server in Asia <> ridiculously long physical path via Europe +- 260-300ms <> Web sQuad JHB; is shorter in BGP than a Google Server in Asia <> peering point to in Hong Kong <> +- 120ms via SAFE <> Web sQuad's network in JHB (3 AS Hops = longer path).

That said, you probably can find a peered speedtest.net server in Hong Kong or Singapore if you look hard enough.