Afrihost CAPPED ADSL Feedback (MTN)

Status
Not open for further replies.
AfriGenie said:
We've been aware of what the exact cause of the shaping glitch is, the issue has largely been actually getting our updates to resolve the glitch, which we are quite confident has now been rectified.

I have to agree with you, we could have handled things differently, and there are quite a few important lessons to be learned from this.
Click to expand...

Good to know. The network seems stable at the moment, holding thumbs it stays that way.
 
Everything running great at the moment. Afrigenie, shoot anyone who touches anything that side now.

233MB app updated in under 5min
 
Last edited:
Davedes said:
This bandwidth issue shouldn't even be a discussion on the capped forum. Uncapped subscribers should be the only accounts experiencing the issues all of us are reporting numerous times s day here with bandwidth being over managed.
Click to expand...
Couldn't agree more. My argument is - as capped clients paying a premium upfront nogal for a product guaranteed never be shaped/throttled AH know exactly how much bandwidth should be available (for capped accounts). Now it seems that, without even notifying us, they use/borrow our already paid for bandwidth to bolster other accounts where people have free reign to use as much as they like. Surely, as this bandwidth is effectively owned by the capped clients, they should not be doing this and seems highly unethical.
Another sore point is that we pay AH for - say 50gigs and whatever we don't use during the month they simply take away from us? This is particularly irksome as some months, due to the bad service, it is impossible to use all that we have paid for. Don't tell me all ISP's do this as they don't. I believe some seem to have a more customer oriented service and allow you to carry your (remember it has been bought and paid for) unused bandwidth over.
I also think that the "Best ISP" results indicate that more clients than AH realise are very unhappy! The Afri-reps sometimes seem to adopt the attitude - Oh well only a few clients are complaining. I am sure though that, like me, when I see someone has raised an issue on this forum, feel that there is no point in me also bogging you down with similar problems. (A bit of laziness is also probably to blame). When I have, they have never been resolved by AH anyway and eventually I just live with it as they eventually go away - only to return later. (Again, just too lazy to change my accounts to another ISP).
Anyway I am the optimistic type and having been with AH for several years live in hope that they may once again get back to the 'glory days' and I can again recommend that friends/family give AH a try.

I have a 10Mbps ADSL line with a 50+50 capped account and am in Cape Town (Southern region).
 
Last edited:
FFS another night not being able to connect to D3, I have 50GBs left on my "capped/unshapped" account but yet again it is USELESS. How long is it going to take to catch the rouge p2p users messing up our connections?
 
Network status is green so there is no issue.

General Notices
Everything is generally hunky dory in the land of Afrihost. Nothing to report, over and out!
 
Ping Test:

ping 197.242.144.102

64 bytes from 197.242.144.102: icmp_req=1 ttl=57 time=23 ms
64 bytes from 197.242.144.102: icmp_req=2 ttl=57 time=27 ms
64 bytes from 197.242.144.102: icmp_req=3 ttl=57 time=25 ms
64 bytes from 197.242.144.102: icmp_req=4 ttl=57 time=24 ms
64 bytes from 197.242.144.102: icmp_req=5 ttl=57 time=22 ms


ping 8.8.8.8

64 bytes from 8.8.8.8: icmp_req=1 ttl=53 time=46 ms
64 bytes from 8.8.8.8: icmp_req=2 ttl=53 time=45 ms
64 bytes from 8.8.8.8: icmp_req=3 ttl=53 time=44 ms
64 bytes from 8.8.8.8: icmp_req=4 ttl=53 time=41 ms
64 bytes from 8.8.8.8: icmp_req=5 ttl=53 time=44 ms


Trace Test:

traceroute -n 197.242.144.102

1 10.0.0.2 1.106 ms * *
2 Request timed out * * *
3 41.181.221.249 73.633 ms 80.763 ms 20.483 ms
4 41.181.221.250 18.19 ms 48.99 ms 19.334 ms
5 41.181.165.115 23.798 ms 21.165 ms 68.193 ms
6 41.181.165.114 17.19 ms 26.668 ms 18.403 ms
7 196.44.0.224 19.107 ms 18.533 ms 20.058 ms
8 196.44.31.99 22.649 ms 21.158 ms 20.961 ms
9 196.30.1.53 23.443 ms 22.125 ms
196.30.1.39 19.336 ms
10 196.31.220.23 21.946 ms
196.31.220.27 20.97 ms 22.081 ms
11 196.31.63.202 19.795 ms
196.31.63.206 19.917 ms 20.562 ms
12 196.30.42.133 21.666 ms 22.396 ms 21.488 ms
13 197.242.144.102 23.7 ms 23.182 ms 22.612 ms
14 undefined NaN ms


traceroute -n 8.8.8.8

1 10.0.0.2 1.365 ms * 1.943 ms
2 Request timed out * * *
3 41.181.221.245 32.951 ms 29.037 ms 23.207 ms
4 41.181.221.246 18.001 ms 17.207 ms 18.363 ms
5 41.181.165.115 20.299 ms 22.569 ms 20.012 ms
6 41.181.165.114 22.794 ms
196.31.180.4 27.016 ms 20.017 ms
7 41.181.180.10 47.142 ms 44.733 ms 43.764 ms
8 196.44.0.43 21.302 ms
196.44.31.170 19.904 ms
196.44.31.94 21.338 ms
9 41.181.139.109 43.892 ms 45.944 ms 44.085 ms
10 72.14.194.74 32.314 ms 31.665 ms 35.075 ms
11 72.14.239.53 22.97 ms
72.14.239.35 20.863 ms
72.14.239.117 19.837 ms
12 8.8.8.8 44.835 ms 44.254 ms 44.342 ms


DNS Test:

nslookup www.afrihost.com

Name: www.afrihost.com
Address: 197.242.144.102


nslookup www.google.com

Name: www.google.com
Address: 216.58.223.36


nslookup thishouldfail.afrihost.com

** server can't find thisshouldfail.com: NXDOMAIN


why would there be 100% loss on the 2nd hop?
 
Davedes said:
Ping Test:

ping 197.242.144.102

64 bytes from 197.242.144.102: icmp_req=1 ttl=57 time=23 ms
64 bytes from 197.242.144.102: icmp_req=2 ttl=57 time=27 ms
64 bytes from 197.242.144.102: icmp_req=3 ttl=57 time=25 ms
64 bytes from 197.242.144.102: icmp_req=4 ttl=57 time=24 ms
64 bytes from 197.242.144.102: icmp_req=5 ttl=57 time=22 ms


ping 8.8.8.8

64 bytes from 8.8.8.8: icmp_req=1 ttl=53 time=46 ms
64 bytes from 8.8.8.8: icmp_req=2 ttl=53 time=45 ms
64 bytes from 8.8.8.8: icmp_req=3 ttl=53 time=44 ms
64 bytes from 8.8.8.8: icmp_req=4 ttl=53 time=41 ms
64 bytes from 8.8.8.8: icmp_req=5 ttl=53 time=44 ms


Trace Test:

traceroute -n 197.242.144.102

1 10.0.0.2 1.106 ms * *
2 Request timed out * * *
3 41.181.221.249 73.633 ms 80.763 ms 20.483 ms
4 41.181.221.250 18.19 ms 48.99 ms 19.334 ms
5 41.181.165.115 23.798 ms 21.165 ms 68.193 ms
6 41.181.165.114 17.19 ms 26.668 ms 18.403 ms
7 196.44.0.224 19.107 ms 18.533 ms 20.058 ms
8 196.44.31.99 22.649 ms 21.158 ms 20.961 ms
9 196.30.1.53 23.443 ms 22.125 ms
196.30.1.39 19.336 ms
10 196.31.220.23 21.946 ms
196.31.220.27 20.97 ms 22.081 ms
11 196.31.63.202 19.795 ms
196.31.63.206 19.917 ms 20.562 ms
12 196.30.42.133 21.666 ms 22.396 ms 21.488 ms
13 197.242.144.102 23.7 ms 23.182 ms 22.612 ms
14 undefined NaN ms


traceroute -n 8.8.8.8

1 10.0.0.2 1.365 ms * 1.943 ms
2 Request timed out * * *
3 41.181.221.245 32.951 ms 29.037 ms 23.207 ms
4 41.181.221.246 18.001 ms 17.207 ms 18.363 ms
5 41.181.165.115 20.299 ms 22.569 ms 20.012 ms
6 41.181.165.114 22.794 ms
196.31.180.4 27.016 ms 20.017 ms
7 41.181.180.10 47.142 ms 44.733 ms 43.764 ms
8 196.44.0.43 21.302 ms
196.44.31.170 19.904 ms
196.44.31.94 21.338 ms
9 41.181.139.109 43.892 ms 45.944 ms 44.085 ms
10 72.14.194.74 32.314 ms 31.665 ms 35.075 ms
11 72.14.239.53 22.97 ms
72.14.239.35 20.863 ms
72.14.239.117 19.837 ms
12 8.8.8.8 44.835 ms 44.254 ms 44.342 ms


DNS Test:

nslookup www.afrihost.com

Name: www.afrihost.com
Address: 197.242.144.102


nslookup www.google.com

Name: www.google.com
Address: 216.58.223.36


nslookup thishouldfail.afrihost.com

** server can't find thisshouldfail.com: NXDOMAIN


why would there be 100% loss on the 2nd hop?
Click to expand...

2nd hop is your exchange. Some exhanges were upgraded recently and are no longer returning pings.
 
Speeds now properly borked, esp intl. Edit: seems only google related services still.
 
Last edited:
Morning Guys

Firstly, thanks for reporting the possible issue with Google services, we have been looking at that and I believe it may have been a routing issue, but we are still waiting for confirmation. So far it seems that services are working correctly, but if anyone is still seeing the same problem, please let me know so we can test with you.

I also hear everyone's arguments here about a capped premium service and I truly understand where you are coming from. I use a capped account and it's pretty frustrating that I can't get full download speed in the evenings at the moment (not always, there have been some good nights recently). I just want to say I know what you are experiencing and I am the first one in the technical team's face asking when this is going to improve. To their credit, despite all the pressure, the guys are really putting in an amazing effort and still have time to try to explain the situation. I do believe that every means available to them is being used to fix this, and they are as committed and as passionate about giving our clients a good experience as we are here on MyBB about getting the right service and value for the money you pay for your product.

When it comes to communication, that's basically our job here, so I'll really try to get more confirmed information wherever I can to report to the forum so you are involved in the process and are up to date.
 
NeoNarchy said:
FFS another night not being able to connect to D3, I have 50GBs left on my "capped/unshapped" account but yet again it is USELESS. How long is it going to take to catch the rouge p2p users messing up our connections?
Click to expand...

Our team has been testing on D3 and we are still able to connect and game normally, so we need to replicate the issue you are experiencing. All Blizzard games use the same traffic signature so other games like Heart of the Swarm, StarCraft would also be suffering, which I don't believe they are. Also that is only in terms of shaping and we definitely don't touch gaming traffic on any product (even shaped).

I can't think off the bat why you're experiencing this issue, I know a few people have reported similar. We'll need to figure out what the common factors are here that links your experience. We've tested D3 on Uncapped and Capped accounts and it appears to be running as expected :(
 
The question is WHEN are you testing these things?

The common factor is that people are having problems from 19:00 till 22:00 in peak time. Testing should take place then as well as from the region the person is from. In addition it should be done on a stocks standard adsl line - not an office fibre connection.

No point in testing in the day. Won't tell you a damn thing :-/
 
Orihalcon said:
The question is WHEN are you testing these things?

The common factor is that people are having problems from 19:00 till 22:00 in peak time. Testing should take place then as well as from the region the person is from. In addition it should be done on a stocks standard adsl line - not an office fibre connection.

No point in testing in the day. Won't tell you a damn thing :-/
Click to expand...

We running reports 24/7 so we're trying to pick up any trends that could help us manage the traffic better and we have picked up many discrepancies already that has helped us improve overall performance. We are definitely looking at the peak periods, but just remember that we have to go back and analyse that data (which is a lot of information) then develop a solution, test it and then implement and monitor.
 
AfriMan said:
We running reports 24/7 so we're trying to pick up any trends that could help us manage the traffic better and we have picked up many discrepancies already that has helped us improve overall performance. We are definitely looking at the peak periods, but just remember that we have to go back and analyse that data (which is a lot of information) then develop a solution, test it and then implement and monitor.
Click to expand...

I was specifically commenting on the gaming problems people are having. To properly replicate those games should be played in the evening from the same region on a stock standard home line. Anything else would be ridiculous to assume you are replicating the test in any measure or form.
 
AfriMan said:
Our team has been testing on D3 and we are still able to connect and game normally, so we need to replicate the issue you are experiencing. All Blizzard games use the same traffic signature so other games like Heart of the Swarm, StarCraft would also be suffering, which I don't believe they are. Also that is only in terms of shaping and we definitely don't touch gaming traffic on any product (even shaped).

I can't think off the bat why you're experiencing this issue, I know a few people have reported similar. We'll need to figure out what the common factors are here that links your experience. We've tested D3 on Uncapped and Capped accounts and it appears to be running as expected :(
Click to expand...

As a matter of interest, where are your team connecting from? Maybe an idea to establish which regions are having issues connecting from the users end.
 
Orihalcon said:
I was specifically commenting on the gaming problems people are having. To properly replicate those games should be played in the evening from the same region on a stock standard home line. Anything else would be ridiculous to assume you are replicating the test in any measure or form.
Click to expand...

As I mentioned, we test across the board, including periods where issues are reported, because we must do empirical testing to ensure that we fully understand the problem. Testing from a single line is also problematic, each home network is unique so we test with multiple lines and multiple clients (especially from the gaming industry where possible) to get a broader picture.
 
BrendanMc said:
As a matter of interest, where are your team connecting from? Maybe an idea to establish which regions are having issues connecting from the users end.
Click to expand...

We test from our offices (on a DSL line) as well as from their homes and we also test with our team members in Durban and Cape Town. We also test with clients in different regions and different line speeds as well as Capped and Uncapped accounts.
 
Status
Not open for further replies.
Back
Top
Sign up to the MyBroadband newsletter
X