Afrihost New Network Feedback - Part 2

Status
Not open for further replies.
AfriMan said:
Johannesburg should be totally unaffected. This is specific to the South.
Click to expand...

No, you misunderstand. I am in Cape Town and using an Afrihost ADSL account to connect with no problem to a network that uses IS fibre in JHB. In fact, I'm tunneling via a VPN client to a VPN server on that network and happily using the internet. So I assume you have an interconnect to IS that is independent of the IPC which is why I'm curious as to where the fibre break occurred.
 
garp said:
No, you misunderstand. I am in Cape Town and using an Afrihost ADSL account to connect with no problem to a network that uses IS fibre in JHB. In fact, I'm tunneling via a VPN client to a VPN server on that network and happily using the internet. So I assume you have an interconnect to IS that is independent of the IPC which is why I'm curious as to where the fibre break occurred.
Click to expand...

I see what you mean now. It's possible that it's routing somewhere before the break, I don't have exact details. But we peer with all local networks at local NAP sites, so that is very possible.
 
I cancelled my uncapped account before Christmas and then last night I changed my mind and enabled it again. Today it isn't working at all and it says I'm capped . I'm in jhb so not sure if this is related to the current situation. Afriman any idea what's going on ? It seems to have gone off at 1am.
 
Pity about the break in the cable, but thanks for keeping us informed. Happy New Year everyone!
 
jvk said:
I cancelled my uncapped account before Christmas and then last night I changed my mind and enabled it again. Today it isn't working at all and it says I'm capped . I'm in jhb so not sure if this is related to the current situation. Afriman any idea what's going on ? It seems to have gone off at 1am.
Click to expand...

It doesn't sound like it. It sounds like we didn't create a new data link for Jan. PM me and I'll fix that for you :)
 
But why you not reply to the ticket i put in this morning 8:00 again your support system not working and no backup after the last fibre brake last year you promise this not happen again 4 1/2 hr
 
AfriMan said:
Rerouting traffic would affect other users, and cause increased latency and congestion, causing us to shape harder. Given the nature of the situation, we need to determine the best course of action for all our clients, including those currently unaffected. We believe the fibre break should be quickly restored, which will instantly restore connectivity, and that this is the fastest and simplest solution for all our clients.
Click to expand...

Shape harder. Great. Everything is still dead for me though. 0% throughput.
And quickly restored. Its now 12:45pm, it's been near 5 hours. I am burning through mobile data thanks to you, will I be compensated, probably not.
 
kohlhtor said:
But why you not reply to the ticket i put in this morning 8:00 again your support system not working and no backup after the last fibre brake last year you promise this not happen again 4 1/2 hr
Click to expand...

I'm really struggling to understand what you are asking here ....

No ISP can promise 100% uptime, a DSL network (or any broadband network) has several thousand moving parts, components, varibales, etc. While we can put redundancies in place for most major issues (that one can expect would require this) it's impossible to run a dual network without running at a massive cost (which clients will have to foot the bill for). There are going to be some times where no amount of planning could prevent some outages. We simply aim to have as much flexibility and options to implement workarounds at different points in the signal chain.

If you submitted a ticket this morning at 8am, i doubt that four hours is a terrible turnaround time, given that there is an outage happening at the moment. You will receive a reply as soon as a consultant is available.
 
Johnatan56 said:
Shape harder. Great. Everything is still dead for me though. 0% throughput.
And quickly restored. Its now 12:45pm, it's been near 5 hours. I am burning through mobile data thanks to you, will I be compensated, probably not.
Click to expand...

Unfortunately we can't compensate in that regard, particularly for a general outage.

What I meant by shape harder is that if we rerouted cape town traffic via johannesburg, we'd have to shape/throttle to the max to try to accomodate the users on the available capacity, and even then it would probably introduce latency. We can do this as a last resort, but it would be best to avoid unless absolutely necessary.
 
AfriMan said:
Unfortunately we can't compensate in that regard, particularly for a general outage.

What I meant by shape harder is that if we rerouted cape town traffic via johannesburg, we'd have to shape/throttle to the max to try to accomodate the users on the available capacity, and even then it would probably introduce latency. We can do this as a last resort, but it would be best to avoid unless absolutely necessary.
Click to expand...

But isn't this the entire reason why you have different tiered/priced packages? So that when something like this happens the people with the cheaper packages get shaped a lot so that the people who are on higher tiered/priced packages can at least load a webpage?
 
Waylander said:
If that is true - it means that capped customers will only be shaped if something like a network fault/failure occurs? In the past, when cables broke or iOS updates ruined the network, Afrihost was quite quick to just tell their customers to suck it up because it was out of their control. So in actual fact, looking at Afrihost's history if a such an event were to happen, everyone would be impacted, even unshaped capped customers.

This seems like a stunt to push people into giving up half their cap, or paying more to keep their cap under a different label of the same ****?

EDIT:

Oh, I now recall that Afrihost stated that the accounts were unshaped, but were under heavy contention due to the network fault and since the new capped unshaped accounts are not exempt from contention build-up they will be impacted just like the shaped accounts.
Click to expand...

kek
 
AfriFella said:
Well, you're right. If a cataclysmic event were to happen again, such as you mentioned above, everyone will be affected. This has got nothing to do with shaping, though. Not sure how you are comparing general network faults or "potential" upstream infrastructure issues (such as undersea cables) to our AUP?

The bottom line is we have spared no expense building and maintaining our DSL network and have constantly been upgrading capacity and network management tools to provide the best experience for our users. There have been some initial teething problems, as everyone knows, but the network is running well and we are now able to offer a wider range of products tailored to compete in the market.
Click to expand...

kekek
 
AfriMan said:
Unfortunately we can't compensate in that regard, particularly for a general outage.

What I meant by shape harder is that if we rerouted cape town traffic via johannesburg, we'd have to shape/throttle to the max to try to accomodate the users on the available capacity, and even then it would probably introduce latency. We can do this as a last resort, but it would be best to avoid unless absolutely necessary.
Click to expand...

I have literally 0% throughout unable to use my net that I pay for, yet it is impossible t reroute traffic as it might inconvenience others. Is that not the entire point I am paying for the prioritized capped account?
Don't you find it hilarious that this happens on the day that the separation of capped accounts occurs, that you shape "shaped capped accounts" under "extreme network load"?

How is a single fiber cable break enough to kill off the entire southern and northern region. It would be understandable to have reduced throughout due to two or more redundant routes for a connection shared by a huge number of your customers, it's got to be one of the dumbest network topologies If only a single cable would bring down half the country.

International cables have redundancy, Seacom and WACS have agreements with each other for this, a minimum of 2 ways out the country extremely geographically separated, how can you have a single cable destroy the entire network for over 5 hours.

Also 3 hours to diagnose this when you should have been able to diagnose it near instantly as the switch before replies, the switch after doesn't so you'd ping from the other side as well and know that the fiber cable is damaged, should take you less than 5 minutes.

This can solely be laid at your feet and easily been avoided, you even have the possibility to reroute via Johannesburg and don't take that options so that your clients can even have a net connection for at least email.

Atrocious networking and decision making.
 
Internet is working now :) .Getting half my line speed but will do a reset later. Cape Town
 
Status
Not open for further replies.
Back
Top
Sign up to the MyBroadband newsletter
X