Web Squad ISP

Status
Not open for further replies.
websquadza said:
Google Update: Google's local CDN isn't serving IPv4 content at the moment, so we are reliant on local caches and transit paths for this traffic. The local caches are serving a large number of ISPs and are under significant pressure.

This is a good time to turn on your IPv6 if you're on a PPPOE connection. Google is still serving IPv6 traffic from their CDN. Most routers support this by default. Log in and enable IPv6 on the PPPOE connection.
Click to expand...
Hello there @websquadza

Appreciate all the feedback on this thread regarding this issue - In the future, would it please be possible to send out either an SMS or email when there are service degradations like these? There is no information about this issue on your Server Status page, and the only reason I learned something was up with international traffic (e.g Youtube), was because I had to look up this thread on MyBB.

Thanks!
 
RynOber771 said:
Hello there @websquadza

Appreciate all the feedback on this thread regarding this issue - In the future, would it please be possible to send out either an SMS or email when there are service degradations like these? There is no information about this issue on your Server Status page, and the only reason I learned something was up with international traffic (e.g Youtube), was because I had to look up this thread on MyBB.

Thanks!
Click to expand...

Thanks for the feedback and we’ll definitely look into this in the future. We do try communicate major outages to our clients quickly- and MyBroadband is a very effective and quick tool for just that. It allows us to gauge the severity and scope of an issue and quickly provide interactive feedback. We’ll definitely look at other channels like sms and email too.

Given the lack of feedback from Vumatel during today’s outage, we couldn’t immediately post an update on our status page- we needed to complete our own investigation first. This was compounded by the sporadic nature of tonight’s outage, affecting areas and regions at random. The first update from Vuma was received around 22:20. This means we need to evaluate the flood of tickets, see if they’re common and then update our network status page accordingly. This we did at 21:30 this evening- the notice is still up until Vumatel confirms the issue is 100% resolved.
 
Last edited:
What is going on today, all day Twitch buffering every few minutes and Netflix image quality is poor
 
How is this still happening,

I'm still being punished by the upstream provider...

diesal@athena:~$ dig news.tweaknews.eu

; <<>> DiG 9.16.27-Debian <<>> news.tweaknews.eu
;; global options: +cmd
;; connection timed out; no servers could be reached
Click to expand...
Edit: maybe I jumped the gun a bit :unsure:
Will continue to monitor.
 
Last edited:
Zodiak said:
What is going on today, all day Twitch buffering every few minutes and Netflix image quality is poor
Click to expand...
Have you checked for last mile packet loss? Netflix caches are local and directly attached- the only reason you’ll struggle with netflix is packet loss. Unfortunately we can’t police every FNO and their every area for PL and you would need to check and let us know. Twitch is very susceptible to any issues unfortunately.
 
Zodiak said:
What is going on today, all day Twitch buffering every few minutes and Netflix image quality is poor
Click to expand...
Netflix was fine for me but Twitch was a complete mess. Not even the lowest setting would work. Worked flawlessly when I activated cloudflare warp.
 
DarkSt0rm said:
View attachment 1274732

View attachment 1274734
View attachment 1274736
View attachment 1274738
View attachment 1274740
Click to expand...
topdog said:
Am also seeing massive packet lost on cogento's network from joburg to overseas destinations. The packet loss carries through from be2355.ccr51.jnb01.atlas.cogentco.com onwards @websquadza
Click to expand...

Multiple NLD failures (CPT<>JHB and JHB<>KZN) are affecting certain upstreams, we have shifted as much traffic as possible to other paths, but there may be some issues that persist.

Update 13:30: Affected provider's NLD path restored
 
Last edited:
We keep talking about NLD and I've received some enquiries with regards to this topic in the past few weeks. SA has multiple National Long Distance paths between key metros, and Web sQuad has capacity of our own on all major providers - linking major metros in a ring and providing resiliency and redundancy for our clients. These NLD paths are major undertakings and are extremely costly. Most of these paths are trenched.

These routes are key to the continued operation of all connectivity in South Africa. Isando has one of the largest exchanges in the world, with over 450+ inter connected networks. These networks all connect to the rest of the world via these NLD cables between KZN, CPT and JHB - because the rest of the world connects in Mtunzini, Ysterfontein, Amanzimtoti (amongst other submarine landing sites).

These routes have also been subject to extensive breaks in the past few months (more than usual) which affect all networks relying on them, to a lesser or greater extent depending on the spread of the breaks and the redundancy measures implemented. These breaks are sometimes accidental (roadworks and construction hitting fibre ducts) and sometimes malicious (attempts at cable theft or plain vandalism)

This morning saw a dual outage on CPT-JHB paths that resulted in certain routes being affected more than others. Our own NLDs stayed up, but some of our upstreams were affected. One route that experienced downtime had been vandalised extensively and I thought I would share the kinds of damage Fibre Operators need to deal with on a daily basis:
2022-03-22 13.11.38.jpg
Photo credit - ZANOG, G Beneke
- Unverified picture of vandalism at a DFA manhole in CPT today, affecting an NLD path to JHB 22/03/2022
 
Status
Not open for further replies.
Back
Top
Sign up to the MyBroadband newsletter
X