Afrihost Business Uncapped Feedback - 2

[Napalm]

Ye guys.. i'm having major latency issues..

Tried to play some counterstrike tonight via steam. Pings of 280ms to 350ms on a local server.

All the other people are having normal pings and only me and another dude also on Afrihost getting above 200ms.



Afrihost please sort out the latency issues :

Tracing route to www.mweb.co.za [196.2.63.110]
over a maximum of 30 hops:

1 1 ms 1 ms 1 ms TP-LINK.Home [192.168.1.254]
2 * * * Request timed out.
3 280 ms 295 ms 280 ms 41.181.221.253
4 314 ms 316 ms 308 ms 41.181.221.254
5 306 ms 317 ms 290 ms tb-dca-2.za--qux-c.za.mtnbusiness.net [41.181.19
8.188]
6 275 ms 279 ms 271 ms unc-cpt-1.mtnns.net [196.44.18.8]
7 320 ms 333 ms 337 ms 196.44.31.106
8 277 ms 279 ms 283 ms nl-ha-2.za--rb-cr-1.za-a.mtnns.net [196.44.0.121
]
9 301 ms 309 ms 315 ms 196-28-178-141.adsl.mweb.co.za [196.28.178.141]

10 272 ms 270 ms 267 ms tengig-0-0-0-1-cpt-p-1.mweb.co.za [197.84.7.1]
11 277 ms 275 ms 287 ms vl-11-cpt-hscore-1.mweb.co.za [197.84.5.238]
12 290 ms 295 ms 304 ms 196.28.178.66
13 391 ms 285 ms 289 ms cte-core-sw2.vwol.net [196.41.144.35]
14 303 ms 304 ms 295 ms www.mweb.co.za [196.2.63.110]

Trace complete.

 

[cortexenslaver]

Hi All


We have a 10Mb ADSL line with an Afrihost uncapped business account, which is supposed to be uncapped and unshaped. However over the last few months we have been experiencing very poor latency on the account.

The results of a local speed test on Afrihost:


When comparing this to a Telkom account, you can clearly see this is an Afrihost issue:


It appears as though Afrihost does not have enough bandwidth available for all of their customers, resulting in severe shaping and throttling.

This is completely unacceptable, and if this is not resolved soon we will take the matter further.
if this is an ipc problem sort it out as capetown has a serious problem

A traceroute to the Afrihost DNS shows that the issue is definitely with MTN:



mars

Same crap here, too....
I have had an open ticket for months... have done MORE than my fair share of tests in the past... have proved that there are major issues in Cape Town (8 months and counting). I received no definitive resolution, got no response for over a month, and then got asked to do the SAME tests AGAIN with more test accounts W T F ???

....and FYI, I have NEVER even been given a "token gesture" of a discount at ANY time.... no problem taking my money though.
I refuse to once again spend the whole month on this forum trying to get this sorted out, only to be told "there's no problem with IPC".
STOP OVER-SELLING YOUR NETWORK AFRIHOST!
Maybe AH should hire REAL tech professionals, who actually know how to manage a network... maybe ask Axxess to lend you their's????
Once again, Axxess account follows the exact same route on the MTN network, with the exact same hops and IP's, YET their network seems unaffected...

GO FIGURE.... This now the 3rd time in a year - FIX IT!!!

 

[slvR]

/snip

maybe move to a different ISP?
these afrijoke issues are not going away any time soon

I'd suggest Crystal Web, Web Africa or Mweb even

 

[Fudzy]

Are you guys able to test your line in the clientzone? Mine keeps failing, information from the router indicates that the signal info is all in order.

 

[cortexenslaver]

Are you guys able to test your line in the clientzone? Mine keeps failing, information from the router indicates that the signal info is all in order.

My CZ hasnt worked properly in months... logged tickets, but they dont get answered, no fault resolution, or I get told the problem must be on my side as AH has tested and all is OK

 

[Fudzy]

My CZ hasnt worked properly in months... logged tickets, but they dont get answered, no fault resolution, or I get told the problem must be on my side as AH has tested and all is OK

What is your SNR like?

 

[Fudzy]

				Downstream		Upstream
Line Coding(Trellis):		On 			On 
SNR Margin (dB):		171			290 
Attenuation (dB):		175 			80 
Output Power (dBm):		0 			126 
Attainable Rate (Kbps):		23536 			1296 
Rate (Kbps):			4096 			511 
D (interleaver depth):		1 			1 
Delay (msec):			0.25 			0.24

 

[Bryn]

My connection just started dying. Massive latency, and throughput gives out after a minute or so after rebooting the router. The connection doesn't actually disconnect - it just loses connectivity.

Edit: Seems okay now.

 

[AfriMan]

Morning everyone,

Seems like this thread kinda exploded, so it's probably going to be easier to post a general message here.

We did see some increased latency in the South over the weekend, it seemed to pick up on Friday night. My reports show a similar increase in latency last night, again, it doesn't seem to be there at the moment, but we'll see as demand picks up. Right now the cause is unknown, we are testing with MTN to see whether this is due to something on our network and if we can resolve this quickly.

I see a lot of posts here mentioning IPC congestion and this is really not the case. While the symptoms seen may be similar to what could be experienced, we'd certainly see demand consistently high, and we'd see packet loss and latency even when we're shaping the network to maximum (for Uncapped accounts, not Business and Capped) and we haven't shaped to maximum for several days now, meaning that there is available bandwidth on the network.

 

[AfriMan]

The latency is a bit lower on webafrica however there is still packet loss on both.

Let me change your line from Fastpath to interleaving tomorrow and see if it makes a difference.

I get the feeling you work for another ISP. If we're all playing open cards here, shouldn't you declare that so we all know who we're dealing with?

 

[electrofreeze]

the problem is we have been having these issues since january and most of the time it is after 5 or over the weekend
then you people are not on the forum to answer questions and your support staff is useless
please bear n mind that we have to pay for this service every month

 

[AfriGenie]

the problem is we have been having these issues since january and most of the time it is after 5 or over the weekend
then you people are not on the forum to answer questions and your support staff is useless
please bear n mind that we have to pay for this service every month

Once we have figured what is causing the spikes in latency and in some cases reduced throughput, we should have a better idea of how long the issue has been occurring.

 

[ghostR]

I get the feeling you work for another ISP. If we're all playing open cards here, shouldn't you declare that so we all know who we're dealing with?

Telkom Lines.

 

[Ho3n3r]

I agree 100% with Electrofreeze. When you have a problem over the weekend, the support staff just resets the line, and maybe tells you to restart your router. If that doesn't work, they're stumped.

Do you guys have any plans to get actual trained staff in these positions, or will these glorified teagirls/boys suffice till the end of time?

 

[AfriMan]

Telkom Lines.

So you work for Telkom? Firstly, I think it's really dodgy to be posting in our threads when you're representing another ISP.

Secondly, does Telkom know that you're posting in public forums claiming to have information that you clearly don't, and also misrepresenting both Telkom and Afrihost? That would seem to me to be a serious breach of business to business agreements and confidentiality on several levels here.

Aside from the fact that your info is wrong, it sounds like what you're doing is pretty wrong.

 

[standerr]

Morning AfriMan

I can confirm mine is a lot better than yesterday. We can call it back to normal for now

I think a lot of us jump on the IPC bandwagon, because that is where we see the spike in latency.
On my tracert it was at ipc-recieve-tb-1a.mtnbusiness.co.za where the 321ms spike started.

 

[AfriMan]

I agree 100% with Electrofreeze. When you have a problem over the weekend, the support staff just resets the line, and maybe tells you to restart your router. If that doesn't work, they're stumped.

Do you guys have any plans to get actual trained staff in these positions, or will these glorified teagirls/boys suffice till the end of time?

We have improved our training, our inductees now go through about 4 weeks intense training and then a probationary period where they are monitored and reviewed on a daily basis. We've also introduced many new software improvements to ensure that our team have more sophisticated tools available.

I totally agree, a port reset is definitely not a fix all solution. However, it does resolve some first line queries so I understand why the team use it extensively. We will get better as our new guys find their feet and gain more experience (with the guidance of our experienced team members)

 

[ghostR]

So you work for Telkom? Firstly, I think it's really dodgy to be posting in our threads when you're representing another ISP.

Secondly, does Telkom know that you're posting in public forums claiming to have information that you clearly don't, and also misrepresenting both Telkom and Afrihost? That would seem to me to be a serious breach of business to business agreements and confidentiality on several levels here.

Aside from the fact that your info is wrong, it sounds like what you're doing is pretty wrong.

I am not representing another ISP. I work for telkom in the lines department and NOT the Telkom Internet Department.

 

[AfriMan]

Morning AfriMan

I can confirm mine is a lot better than yesterday. We can call it back to normal for now

I think a lot of us jump on the IPC bandwagon, because that is where we see the spike in latency.
On my tracert it was at ipc-recieve-tb-1a.mtnbusiness.co.za where the 321ms spike started.

From the reports it did seem that the latency was internal, but we haven't isolated particular cause. Any link in the signal chain could induce latency, so we have to test from end to end before we can make a conclusion.

Glad things are better for you though

 

[Ho3n3r]

We have improved our training, our inductees now go through about 4 weeks intense training and then a probationary period where they are monitored and reviewed on a daily basis. We've also introduced many new software improvements to ensure that our team have more sophisticated tools available.

I totally agree, a port reset is definitely not a fix all solution. However, it does resolve some first line queries so I understand why the team use it extensively. We will get better as our new guys find their feet and gain more experience (with the guidance of our experienced team members)

That's why I mentioned that they have no clue what to do after this is tried, and fails.