Afrihost Capped ADSL Feedback (MTN) - New beginning

Status
Not open for further replies.
The stream was only 56K on a 10Meg line so hardly a noticeable. It was fine all day as you say but at 10 PM it started cutting off. Not just the stream but everything, couldn't use Google etc. rebooting the router was the only way to get the connection back.
 
AfriMan said:
We're definitely aware of the rise in latency that we see over the peak demand period in the evenings. Some evenings have definitely been better, but when demand is excessive we do see latency rise and some services very dependant on low latency like gaming could show some packet loss. The main thing is that this is only the case during that peak demand window. When demand is normal, then we should see great latency as well as zero packet loss (at least from our network).
Click to expand...

Yes the internet doesn't work during the peak period, when most people need to use it, that's the main thing. The silver lining if you will.
 
Unhappy438 said:
Yes the internet doesn't work during the peak period, when most people need to use it, that's the main thing. The silver lining if you will.
Click to expand...

Yes, they are effectively saying "Our network work's perfectly as long as you don't try to use it".
 
Having the same problem here in PE. Works fine at 6 o'clock in the morning but then seems to get progressively worse as the day wears on. That makes it sound like a BANDWIDTH problem AFRIHOST. I'm ready to jump ship, and will pay more for a better service.
 
Prodegy said:
So let me get this right.

I use my internet at home in the evenings and because its PEAK time I cant expect a usable service for gaming on your network because of demand? I only use the account in the evenings FOR gaming so afrihost isnt the service for that then? even though it used to be for years? In the day im sure its fine I raid at night in WoW I work in the day so then 100% of your service isnt usable for me because I need it during "peak" hours? So gamers in the south because of demand must just move elsewhere or just wait till 12:00 when the peak is over to see no packet loss?

Is it just me or is that just sound like the worst service in the world?
Click to expand...

U know I would love an answer here, what exactly should I do in the evenings to game besides PAY another ISP?
 
Can anyone tell me if this is normal?

Tracing route to www.youtube.com [83.170.83.224]
over a maximum of 30 hops:

1 3 ms 10 ms 3 ms HOME [192.168.1.1]
2 14 ms 10 ms 11 ms 105-236-14-1-esr-lo.mtnbusiness.co.za [105.236.14.1]
3 20 ms 19 ms 20 ms 41.181.221.218
4 22 ms 20 ms 21 ms rb-dca-2.za--qux-b.za.mtnbusiness.net [41.181.165.118]
5 218 ms 217 ms 217 ms rb-cr-2.za--rb-dca-2.za-b.mtnns.net [196.44.0.155]
6 218 ms 225 ms 217 ms jh-cr-1.za--rb-cr-2.za.mtnns.net [196.44.0.42]
7 217 ms 230 ms 233 ms ls-cr-2.uk--jh-cr-1.za.mtnns.net [196.44.31.114]
8 217 ms 218 ms 219 ms ls-pr-2.uk--ls-cr-2.uk-a.mtn.net [209.212.111.187]
9 221 ms 219 ms 219 ms linx-226.as13213.net [195.66.236.19]
10 219 ms 218 ms 217 ms 83.170.70.230
11 221 ms 219 ms 219 ms dc5.as13213.net [83.170.70.134]
12 229 ms 226 ms 232 ms vps1.helpmego.to [83.170.83.224]

Trace complete.
 
Why are you going to the international cache? Did you type the IP instead of the address?
 
To be honest, I posted the tracert because youtube was feeling a little sluggish, and I noticed all the +200ms hops. I didn't even realize that it was resolving www.youtube.com to a weird address..

Here's a tracert using an IS-based test account, using 'www.youtube.com' and still using UnoTelly:

Tracing route to youtube-ui.l.google.com [64.233.184.190]
over a maximum of 30 hops:

1 5 ms 2 ms 6 ms HOME [192.168.1.1]
2 10 ms 13 ms 10 ms 196-215-220-1.dynamic.isadsl.co.za [196.215.220.1]
3 12 ms 14 ms 19 ms cdsl1-umh-gi0-0-0-3102.ip.isnet.net [196.38.74.245]
4 13 ms 13 ms 12 ms 196.26.210.200
5 13 ms 13 ms 15 ms mi-za-umh-p8-ten-0-0-0-1.ip.isnet.net [168.209.93.240]
6 29 ms 26 ms 27 ms 168.209.0.178
7 30 ms 29 ms 29 ms core1a-bry-gi0-0-0.ip.isnet.net [168.209.217.1]
8 25 ms 25 ms 26 ms core1a-bry-ge1-0-1.isnet.net [168.209.100.237]
9 25 ms 25 ms 25 ms 196.26.0.130
10 25 ms 29 ms 25 ms 72.14.205.16
11 198 ms 198 ms 197 ms 66.249.95.8
12 194 ms 192 ms 194 ms proxy.google.com [216.239.51.5]
13 205 ms 203 ms 206 ms 64.233.174.82
14 198 ms 198 ms 198 ms 209.85.255.223
15 * * * Request timed out.
16 203 ms 203 ms 207 ms 64.233.184.190

Trace complete.
 
Ok nevermind, I've switched back to my Afrihost capped account and now it's resolving properly, and Youtube is behaving normally.

Still, I'd love to know wtf was happening just now, lol :D
 
AfriMan said:
We're definitely aware of the rise in latency that we see over the peak demand period in the evenings. Some evenings have definitely been better, but when demand is excessive we do see latency rise and some services very dependant on low latency like gaming could show some packet loss. The main thing is that this is only the case during that peak demand window. When demand is normal, then we should see great latency as well as zero packet loss (at least from our network).
Click to expand...

Please tell me you are joking with this reply? Absolutely ludicrous!
 
4415711870.png


4415703492.png
 
Morning All

I see most of the posts relate to speeds or performance in the South, and some comments I made in a post. I think it's best to address this in one post, so that there is no confusion.

We've mentioned several times that the issue we are currently dealing with is that demand is currently high, especially in the evenings, when we see more P2P traffic than we'd like and this additional demand places strain on the available network resources. We've improved our deep packet inspection accuracy, which drives our Uncapped shaping, which has significantly reduced demand and we are seeing more and more evenings when latency is not affected and where it does the overall affect is greatly reduced. There is still more work to be done until we can reduce the excess traffic to the point where performance is within the expected levels at all times. Right now, we're seeing good performance in terms of latency throughout the day except for a peak window somewhere between 9pm and 10pm and we are focussing our efforts on making clients internet experience as good as it can be as we continue to work on reducing demand from rogue P2P traffic.

There seems to be a feeling that we're not saying we don't have enough capacity. If all traffic was unshaped on all products, yes, we would not have sufficient capacity to deliver a good experience. However, we do believe that with shaping working at optimum levels and all traffic being correctly classified, we believe that have adequate resources to deliver a good overall experience. If shaping and packet inspection is brought to full efficiency and we're still not able to provide the experience we expect, then we'll obviously need to consider other options in the way of capacity or a different approach to managing the network. For now though, from our data, we believe that we can meet demand with these elements under control.
 
prodigyX said:
Ok nevermind, I've switched back to my Afrihost capped account and now it's resolving properly, and Youtube is behaving normally.

Still, I'd love to know wtf was happening just now, lol :D
Click to expand...

we have seen a few problems with YouTube in the last week or so, I'm not sure if this is related. My own experience over the weekend seemed pretty normal. We'll test more extensively on our side as well just to be sure all is working as it should.
 
Status
Not open for further replies.
Back
Top
Sign up to the MyBroadband newsletter
X