Afrihost Uncapped ADSL Feedback (Pt3)

Status
Not open for further replies.
HOA-842-69189 / GZR-358-71829
Please liaise with AfriGenie. I want this ticket taken away from whoever the incompetent was that decided that the way to "solve" this was to enable shaping. See last response on HOA-842-69189 ticket for name etc.

I am asked to run tracerts and do debugging on the account all day long and this incompetent responds 2am in the morning to me with and I quote:
Thank you for your mail, I trust you are well.
[...]
I have taken a look into your account and have now activated your Business optimization, this will slow down your downloads but real time protocols should speed up.
[...etc other bull]

No I am not well. Is this the kind of support you offer your clients on business accounts? I had been debugging with one technical person that "appeared" to know whats going on only to out of the blue get some junior grad school techie thumb suck a answer 2am in the morning and decide that I would prefer shaping on my account?

Now, I want you to tell me when you are going to fix this routing/congestion problem in CT, and I want to know by this afternoon (3pm - mail me I will be in meetings most of the day). I will either know the time frame or I will no longer be your client, everyone I know will no longer be your client, and every person I meet or talk to will no longer and never be your client. As you can tell probably this is the end of the road between me and you, either you deliver service, or I will find myself a provider who can actually deliver.

Here are all the stats again.
DATA:
3am
=Afrihost=
Tracing route to www.afrihost.com [41.181.185.18]
over a maximum of 30 hops:

1 1 ms 1 ms 1 ms 192.168.0.1
2 12 ms 11 ms 11 ms 105-236-5-193-esr-lo.mtnbusiness.co.za [105.236.5.193]
3 12 ms 13 ms 13 ms ipc-recieve-tb-2a.mtnbusiness.net [41.181.54.86]

4 12 ms 12 ms 12 ms tb-dca-2.za--qux-c.za.mtnbusiness.net [41.181.198.188]
5 13 ms 13 ms 13 ms unc-cpt-1.mtnns.net [196.44.18.8]
6 33 ms 33 ms 33 ms pe-cr-2.za--el-mse-1.za.mtnns.net [196.44.31.66]

7 31 ms 32 ms 33 ms jh-dca-1.za--jh-cr-1.za-a.mtnns.net [196.44.0.217]
8 34 ms 32 ms 33 ms 18.185.181.41.in-addr.aserv.co.za [41.181.185.18]
good.

11:45am
=Afrihost=
Tracing route to www.afrihost.com [41.181.185.18]
over a maximum of 30 hops:

1 <1 ms <1 ms <1 ms 192.168.0.1
2 11 ms 10 ms 10 ms 105-236-5-193-esr-lo.mtnbusiness.co.za [105.236.5.193]
3 115 ms 113 ms 111 ms ipc-recieve-tb-2a.mtnbusiness.net [41.181.54.86]

4 123 ms 117 ms 115 ms tb-dca-2.za--qux-c.za.mtnbusiness.net [41.181.198.188]
5 114 ms 115 ms 120 ms unc-cpt-1.mtnns.net [196.44.18.8]
6 118 ms 124 ms 122 ms jh-cr-1.za--tb-cr-2.za.mtnns.net [196.44.31.66]
7 181 ms 176 ms 161 ms jh-dca-1.za--jh-cr-1.za-a.mtnns.net [196.44.0.217]
8 137 ms 126 ms 123 ms 18.185.181.41.in-addr.aserv.co.za [41.181.185.18]

Trace complete.
bad.

=Web Africa=

Tracing route to www.afrihost.com [41.181.185.18]
over a maximum of 30 hops:

1 <1 ms 1 ms 1 ms 192.168.0.1
2 12 ms 11 ms 11 ms 196-210-150-129.dynamic.isadsl.co.za [196.210.150.129]
3 13 ms 13 ms 12 ms cdsl2-ctn-vl2276.ip.isnet.net [196.38.72.125]
4 12 ms 13 ms 12 ms 196.35.115.136
5 13 ms 13 ms 13 ms mi-za-cpt-p8-te0-0-0-0.ip.isnet.net [168.209.6.13]
6 12 ms 12 ms 12 ms 168.209.6.130

7 12 ms 12 ms 12 ms 41-208-29-210.mtnns.net [41.208.29.210]
8 13 ms 16 ms 13 ms rb-cr-1.za.nl-ha-2.za-a.mtnns.net [196.44.0.120]
9 14 ms 14 ms 14 ms tb-cr-2.za--ct-cr-1.za-a.mtnns.net [196.44.31.107]
10 32 ms 32 ms 32 ms jh-cr-1.za--tb-cr-2.za.mtnns.net [196.44.31.66]
11 34 ms 35 ms 35 ms jh-dca-1.za--jh-cr-1.za-a.mtnns.net [196.44.0.217]
12 34 ms 34 ms 35 ms 18.185.181.41.in-addr.aserv.co.za [41.181.185.18]

Trace complete.
good.

=5pm=
=Afrihost=

Tracing route to www.afrihost.co.za [41.181.185.18]
over a maximum of 30 hops:

1 1 ms 1 ms 1 ms 192.168.0.1
2 11 ms 11 ms 11 ms 105-236-5-193-esr-lo.mtnbusiness.co.za [105.236.5.193]
3 356 ms 341 ms 340 ms ipc-recieve-tb-2a.mtnbusiness.net [41.181.54.86]

4 392 ms 367 ms 363 ms tb-dca-2.za--qux-c.za.mtnbusiness.net [41.181.198.188]
5 334 ms 335 ms 333 ms unc-cpt-1.mtnns.net [196.44.18.8]
6 387 ms 374 ms 379 ms pe-cr-2.za--el-mse-1.za.mtnns.net [196.44.31.66]
7 397 ms 368 ms 384 ms jh-dca-1.za--jh-cr-1.za-a.mtnns.net [196.44.0.217]
8 366 ms 370 ms 373 ms 18.185.181.41.in-addr.aserv.co.za [41.181.185.18]

Trace complete.
bad.


=WebAfrica=

Tracing route to www.afrihost.co.za [41.181.185.18]
over a maximum of 30 hops:

1 <1 ms 1 ms 1 ms 192.168.0.1
2 12 ms 11 ms 11 ms 196-210-150-129.dynamic.isadsl.co.za [196.210.150.129]
3 12 ms 12 ms 12 ms cdsl1-ctn-vl2173.ip.isnet.net [196.38.72.113]
4 12 ms 13 ms 13 ms 196.35.115.128
5 13 ms 13 ms 14 ms mi-za-cpt-p8-te0-7-0-0.ip.isnet.net [168.209.2.13]
6 13 ms 12 ms 12 ms 168.209.6.131

7 13 ms 14 ms 13 ms mtnns-2.cinx.net.za [196.223.22.31]
8 14 ms 13 ms 13 ms rb-cr-1.za.nl-ha-2.za-a.mtnns.net [196.44.0.120]
9 13 ms 14 ms 13 ms tb-cr-2.za--ct-cr-1.za-a.mtnns.net [196.44.31.107]
10 35 ms 35 ms 35 ms pe-cr-2.za--el-mse-1.za.mtnns.net [196.44.31.66]
11 31 ms 31 ms 31 ms jh-dca-1.za--jh-cr-1.za-a.mtnns.net [196.44.0.217]
12 31 ms 32 ms 32 ms 18.185.181.41.in-addr.aserv.co.za [41.181.185.18]

Trace complete.
good.

=8:50pm=
=Afrihost=

Tracing route to www.afrihost.co.za [41.181.185.18]
over a maximum of 30 hops:

1 <1 ms 1 ms 1 ms 192.168.0.1
2 11 ms 11 ms 11 ms 105-236-5-193-esr-lo.mtnbusiness.co.za [105.236.5.193]
3 400 ms 390 ms 387 ms ipc-recieve-tb-2a.mtnbusiness.net [41.181.54.86]

4 404 ms 415 ms 421 ms tb-dca-2.za--qux-c.za.mtnbusiness.net [41.181.198.188]
5 403 ms 395 ms 397 ms unc-cpt-1.mtnns.net [196.44.18.8]
6 396 ms 397 ms 401 ms jh-cr-1.za--tb-cr-2.za.mtnns.net [196.44.31.66]
7 411 ms 451 ms 420 ms jh-dca-1.za--jh-cr-1.za-a.mtnns.net [196.44.0.217]
8 412 ms 418 ms 422 ms 18.185.181.41.in-addr.aserv.co.za [41.181.185.18]

Trace complete.
bad.
=Web Africa=
Tracing route to www.afrihost.co.za [41.181.185.18]
over a maximum of 30 hops:

1 <1 ms 1 ms 1 ms 192.168.0.1
2 11 ms 11 ms 11 ms 196-210-150-129.dynamic.isadsl.co.za [196.210.150.129]
3 12 ms 13 ms 12 ms cdsl2-ctn-vl2276.ip.isnet.net [196.38.72.125]
4 13 ms 12 ms 12 ms 196.35.115.136
5 14 ms 14 ms 13 ms mi-za-cpt-p8-te0-7-0-0.ip.isnet.net [168.209.2.13]

6 13 ms 13 ms 13 ms 168.209.2.130
7 12 ms 46 ms 13 ms 41-208-29-210.mtnns.net [41.208.29.210]
8 13 ms 14 ms 14 ms rb-cr-1.za.nl-ha-2.za-a.mtnns.net [196.44.0.120]
9 15 ms 14 ms 14 ms tb-cr-2.za--ct-cr-1.za-a.mtnns.net [196.44.31.107]
10 35 ms 36 ms 35 ms jh-cr-1.za--tb-cr-2.za.mtnns.net [196.44.31.66]
11 35 ms 35 ms 35 ms jh-dca-1.za--jh-cr-1.za-a.mtnns.net [196.44.0.217]
12 31 ms 32 ms 32 ms 18.185.181.41.in-addr.aserv.co.za [41.181.185.18]

Trace complete.

good.

Enabling Business Priorisation was definitely a poor move on our part, we'll chase up that ticket and have a word with the agent concerned. Really sorry about that :(

We are seeing from reports that latency was poor yesterday. Unfortunately none of the MyBB reps were on duty yesterday, so we are reporting this to the team this morning, but they also run the same reports that we have access to so that are also aware of the problem. We'll report from a user perspective that this was a serious issue in terms of performance :(
 
From the two traces definitely looks like latency on our side. My reports show that there was high demand in the South and East over the public holiday and we definitely saw intermittent latency mostly in the afternoon to early evening. We have reported this to the network team :(
Is this going to be addressed appropriately or just hope the network doesn't spike again? I hope the former.
 
Is this going to be addressed appropriately or just hope the network doesn't spike again? I hope the former.

Definitely!

This has a very high priority at the moment, and we're taking a long hard look at what exactly is going wrong.
 
Dota 2 is completely a waste of time with Afrihost, local servers over 100ms and goes up and down.
 
It bounces around. I at one point was getting amazing pings of < 50ms and for most of the time > 150ms.

AH and axxess are getting terrible reps lately - whenever I paused for ping issues the 1st question that was asked in 3 games was: "AH?".
 
It bounces around. I at one point was getting amazing pings of < 50ms and for most of the time > 150ms.

AH and axxess are getting terrible reps lately - whenever I paused for ping issues the 1st question that was asked in 3 games was: "AH?".

That's definitely worrying in terms of reputation, and it's something we definitely want to make sure we get on top of ASAP. We are closely monitoring ping times on the network and have setup as many early warning systems as we can to try to kill any additional latency whenever and wherever we can :(
 
Afternoon MyBB Guys!

We've been in discussion with the Higher-Ups, and we've now updated our Noticeboard with some details on the current network status and the upgrades that we have planned.

At the moment I can't go into anymore detail, but you're welcome to post any questions that you may have and I'll do my best to answer. :)
 
Hi Everyone,

I have developed a new tool to do a constant ping test and upload the results (just the ping result number) to a central server, so that all Afrihost users will be able to see when the network is experiencing ping-issues :) Hopefully it will help the reps out when they need to know if the network is having issues again. You can view the source code HERE to make sure that nothing personal is uploaded if you'd like.

I need your help to get it going. Could you please download the following java file, and let it run in the background:
http://releases.zertop.com/latencyreport/Releases/LatencyReportV1.0.0.jar

You can view the collected results here:
http://releases.zertop.com/latencyreport/

Thanks!
Z
 
Afternoon MyBB Guys!

We've been in discussion with the Higher-Ups, and we've now updated our Noticeboard with some details on the current network status and the upgrades that we have planned.

At the moment I can't go into anymore detail, but you're welcome to post any questions that you may have and I'll do my best to answer. :)

It is good to see you are acknowledging the issue on the board. And the latency looks good from my side so far today.
8:15pm
3 13 ms 12 ms 12 ms ipc-recieve-tb-1a.mtnbusiness.net
Thanks for your efforts while taking the heat.

I would suggest that there be more support communication though when handing over support tickets from one support person to another: there is nothing that gets your blood boiling like starting over with a new person again and they think its your line that's at fault after you have already gone through the great effort of proving it is not with the prior support person.
 
It is good to see you are acknowledging the issue on the board. And the latency looks good from my side so far today.
8:15pm
3 13 ms 12 ms 12 ms ipc-recieve-tb-1a.mtnbusiness.net
Thanks for your efforts while taking the heat.

I would suggest that there be more support communication though when handing over support tickets from one support person to another: there is nothing that gets your blood boiling like starting over with a new person again and they think its your line that's at fault after you have already gone through the great effort of proving it is not with the prior support person.

Great to hear that things are improving!
Around your suggestion, it's hard to communicate the full issue as tickets are picked up and attended to by different agents. Saying that though, simply reviewing what was previously discussed in the ticket should get everyone on the right track.
 
Great to hear that things are improving!
Around your suggestion, it's hard to communicate the full issue as tickets are picked up and attended to by different agents. Saying that though, simply reviewing what was previously discussed in the ticket should get everyone on the right track.
10:30pm
3 109 ms 109 ms 165 ms ipc-recieve-tb-1a.mtnbusiness.net

Problem is still there ironically it didn't show up during peak.
 
10:30pm
3 109 ms 109 ms 165 ms ipc-recieve-tb-1a.mtnbusiness.net

Problem is still there ironically it didn't show up during peak.

Although they have no fully admitted it, their bandwidth management system is a very large turd, and I would guess they are now pulling it out and replacing it with something else. Here is to hoping whatever is being done, fixes the problem....
 
Afrihost was (and still is as I'm typing) so completely and utterly useless this weekend.
Torrents: 20kb/s
Direct download: 90kb/s
Packet loss: Between 8% and 59%

Switch to other ISP
Torrents: 380kb/s
Direct Download: 411kb/s
Packet loss: 0%

and now I'm just waiting for them to blame my line again or tell me again that packet loss is because of the teamspeak server (even though 4 other ISP's that I've tried work flawlessly on it and worked flawlessly on afrihost for years)
For the Majority of Sunday I couldn't open half the sites that I wanted to. After struggling for about 30 mins to load a website, I finally realized that there wasn't anything wrong with the site, it was just Afrihost that couldn't let me load the damn thing.
 
10:30pm
3 109 ms 109 ms 165 ms ipc-recieve-tb-1a.mtnbusiness.net

Problem is still there ironically it didn't show up during peak.

It depends on demand on the network, which can really be whenever uses logon. The weekend will usually differ from weekdays but we also find that each behaves differently with different peak windows.
 
Although they have no fully admitted it, their bandwidth management system is a very large turd, and I would guess they are now pulling it out and replacing it with something else. Here is to hoping whatever is being done, fixes the problem....

I can't say what the upgrades will entail, that is up to the network team and our CTO who will be reviewing the entire network to see what changes can be made. The system we use is internationally recognised and used by several other ISPs but maybe it's not suitable for the way we'd like to run our network to be as proactive as possible in giving our clients as much bandwidth as possible when any drop in demand allows it.

Will hopefully be able to share more specific details when we have them.
 
I can't say what the upgrades will entail, that is up to the network team and our CTO who will be reviewing the entire network to see what changes can be made. The system we use is internationally recognised and used by several other ISPs but maybe it's not suitable for the way we'd like to run our network to be as proactive as possible in giving our clients as much bandwidth as possible when any drop in demand allows it.

Will hopefully be able to share more specific details when we have them.

Any news on the elusive "card" that needs to be fixed at my "exchange". Surely something has to be done? Heading for 8 Months now, almost a whole pregnancy....
 
Afrihost was (and still is as I'm typing) so completely and utterly useless this weekend.
Torrents: 20kb/s
Direct download: 90kb/s
Packet loss: Between 8% and 59%

Switch to other ISP
Torrents: 380kb/s
Direct Download: 411kb/s
Packet loss: 0%

and now I'm just waiting for them to blame my line again or tell me again that packet loss is because of the teamspeak server (even though 4 other ISP's that I've tried work flawlessly on it and worked flawlessly on afrihost for years)
For the Majority of Sunday I couldn't open half the sites that I wanted to. After struggling for about 30 mins to load a website, I finally realized that there wasn't anything wrong with the site, it was just Afrihost that couldn't let me load the damn thing.

Loading websites should generally be fine, I definitely did not experience those kind of problems on my own connection, however if you are testing from a different region during peak latency, then it would be possible that you would have a different experience. When there is no undue latency on the network from our side, TS definitely works on standard and most non-standard ports. When latency is high, especially no-standard ports suffer, but overall latency dependant services could be poor.
 
I can't say what the upgrades will entail, that is up to the network team and our CTO who will be reviewing the entire network to see what changes can be made. The system we use is internationally recognised and used by several other ISPs but maybe it's not suitable for the way we'd like to run our network to be as proactive as possible in giving our clients as much bandwidth as possible when any drop in demand allows it.

Will hopefully be able to share more specific details when we have them.

The problem here is that when Afrihost deals with network upgrades, they seem to be reactive instead of proactive.

The end-user should not "feel" the upgrade process. Yet here we are. Packet loss, poor throughput, the works going by this thread alone.
 
The problem here is that when Afrihost deals with network upgrades, they seem to be reactive instead of proactive.

The end-user should not "feel" the upgrade process. Yet here we are. Packet loss, poor throughput, the works going by this thread alone.

I'm in definite agreement with you!

Going forward these changes will allow us a great deal more dynamic control over the network, and any further upgrades can be completed rapidly and as needed.
 
Status
Not open for further replies.
Top
Sign up to the MyBroadband newsletter
X