Afrihost Business Uncapped Feedback

Status
Not open for further replies.
Why is it a weird case? It's been proven long ago that AH have capacity issues. You just stringing the guy along and wasting his time.

Its weird because he has over 1000ms pings to his exchange - causing issues! However, although it shows that he has the same on MWEB, he says that it doesnt affect him? I know there are capcity issues, but I don't think this has anything to do with them
 
This used to be such a quiet thread. Quite weird seeing all this troubleshooting.
 
Its weird because he has over 1000ms pings to his exchange - causing issues! However, although it shows that he has the same on MWEB, he says that it doesnt affect him? I know there are capcity issues, but I don't think this has anything to do with them

Zertop ... thanks for all your input.

For me the proof is in the using, and now on Mweb, my Skype is working, my Teamviewer is OK, Cirtix VPN is fine etc.
This weekend Citrix was a nightmare on AH. Constant disconnects. It used to be just fine.

A simple test I run is Twitch.tv. Not a scientific test, but it points to something not right.
Twitch used to work on Afrihost business on medium without buffering. Then something changed, it cannot stream now without buffering every 20-30s. All the time.
I switch to WebAfrica, and it streams smoothly... for as long as I leave it running, it streams on medium, and sometimes on high, without a buffer at all.
I stream on Mweb or Axxess buffer-free on medium for as long as the stream is open.
On Afrihost, it cannot stream on medium without constant buffering. It does not matter what time of day or night, or which stream, it buffers. Constantly.

If there is a line fault or exchange issue, I need this to be resolved. It will only be a matter of time before every ISP becomes schizo like AH is now.
If Telkom report back that all is OK, well then, time to give one of the others a run and see how long they behave for.

Your input and these tests have been very helpful. The last Mweb one shows that there is something amiss.
Perhaps these other ISPs are more fault tolerant than AH?

Thanks again for your time and input.
 
Zertop ... thanks for all your input.

For me the proof is in the using, and now on Mweb, my Skype is working, my Teamviewer is OK, Cirtix VPN is fine etc.
This weekend Citrix was a nightmare on AH. Constant disconnects. It used to be just fine.

A simple test I run is Twitch.tv. Not a scientific test, but it points to something not right.
Twitch used to work on Afrihost business on medium without buffering. Then something changed, it cannot stream now without buffering every 20-30s. All the time.
I switch to WebAfrica, and it streams smoothly... for as long as I leave it running, it streams on medium, and sometimes on high, without a buffer at all.
I stream on Mweb or Axxess buffer-free on medium for as long as the stream is open.
On Afrihost, it cannot stream on medium without constant buffering. It does not matter what time of day or night, or which stream, it buffers. Constantly.

If there is a line fault or exchange issue, I need this to be resolved. It will only be a matter of time before every ISP becomes schizo like AH is now.
If Telkom report back that all is OK, well then, time to give one of the others a run and see how long they behave for.

Your input and these tests have been very helpful. The last Mweb one shows that there is something amiss.
Perhaps these other ISPs are more fault tolerant than AH?

Thanks again for your time and input.

Firstly - no probs :D Always happy to help

Secondly - Yea, i genuinely have no clue. It could be (as Chevron said), a capacity issue. However that doesn't explain the ridiculous pings. I guess if MWEB works for you, thats what you are going to have to use for now until you get some proper answers.

Z
 
Its weird because he has over 1000ms pings to his exchange - causing issues!

You can ping the exchange :wtf: I think you got it wrong, you actually pinging the ESR...the exchange is transparent and will not show up on pings and tracert's

Correct me if I'm wrong.
 
I think I would agree with Zertop ito there is either a line fault or more probably exchange congestion at times.
I have seen several people mentioning that it seems that Telkom are working on lines and it seems to create havoc atm.
Not sure if that is applicable to the WC, maybe more in the rain affected areas?
 
Yea, definitely a line issue - cause thats pinging your exchange, and should be under ~25ms at all times! I think the best thing would be to report a fault :(

That could account for all your issues.

My networking knowledge is obviously not quite where you guys are (although its not bad :whistle: ).

A quick question then.

Why is i.got~issues MWEB Exchange IP 197.84.96.138, but his Afrihost Exchange IP 41.181.53.150. Is the exchange not the grey Telkom box sitting 1-3km away from his place? Why then do different ISP's use different Exchange IP's, if it's the same box?
 
My networking knowledge is obviously not quite where you guys are (although its not bad :whistle: ).

A quick question then.

Why is i.got~issues MWEB Exchange IP 197.84.96.138, but his Afrihost Exchange IP 41.181.53.150. Is the exchange not the grey Telkom box sitting 1-3km away from his place? Why then do different ISP's use different Exchange IP's, if it's the same box?

The little grey box is called a DSLAM. Its pretty much the link between the Exchange and the User. The exchange is a physical building in a town where all the telkom lines come into. The "exchange" per se isn't shown, but rather is the middle link. The second hop of a tracert shows not the exchange, but rather the place just before getting onto the ISP's network.

Therefore - we can use it to calculate whether the problem lies with the ISP, or whether it lies with the line/exchange. Each ISP has a different "handler", which accounts for the different IP's :)

I could be wrong, because my knowledge on this isn't the greatest - but someone else should be able to tell you if that is correct :)
 
I am sort of stuck as to where to go

I have sent a lengthy ticket to Afrihost a couple of weeks ago the performance of my line is kinda bad (ticket DUU-647-86661) nothing has improved, port resets etc

I have two accounts with afrihost 2meg business uncapped and a 10 gb capped account (I have a 4 meg line so I was hoping to get full line speed on the capped account)

To quote my ticket

I'm getting bad speeds on my 2meg business account (note I'm on a 4meg line)

http://www.speedtest.net/my-result/3360308754

Speeds on my 10 GIG Capped account (also AH)

http://www.speedtest.net/my-result/3360354660

tracert to google (on Capped account)

1 1 ms 1 ms 1 ms 10.0.0.2
2 54 ms 53 ms 60 ms 105-236-5-65-esr-lo.mtnbusiness.co.za [105.236.5.65]
3 31 ms 34 ms 45 ms ipc-send-tb-3a.mtnbusiness.net [41.181.53.214]
4 87 ms 96 ms 90 ms tb-dca-2.za--qux-c.za.mtnbusiness.net [41.181.198.188]
5 85 ms 84 ms 68 ms compj-cpt-1.mtnns.net [196.44.18.2]
6 220 ms 247 ms 238 ms ls-cr-2--tb-cr-1.uk-b.mtnns.net [196.44.31.113]
7 249 ms * 207 ms am-cr-1.nl--ls-cr-2.uk-a.mtnns.net [196.44.31.183]
8 196 ms 186 ms 211 ms am-tpr-1.nl--am-cr-1.nl-a.mtn.net [209.212.111.141]
9 223 ms 220 ms 216 ms core2.ams.net.google.com [195.69.145.100]
10 263 ms 319 ms * 209.85.254.90
11 227 ms 178 ms 223 ms 72.14.238.153
12 231 ms 215 ms * 209.85.240.29
13 221 ms 226 ms 218 ms 209.85.244.241
14 225 ms 207 ms 225 ms 209.85.253.31
15 * 218 ms 243 ms lhr14s19-in-f5.1e100.net [173.194.34.69]

Trace complete.
tracert to google capped account

Tracing route to google.com [173.194.34.98]
over a maximum of 30 hops:

Nothing has changed, the tracert is basically the same and so are general pings to services. YouTube buffers like hell

I have gotten port reset after port reset

Line Stats : http://imgur.com/d91Mi5G

What I dont get is I have a high attainable rate but my line is worse than a fish right now. Speedtests dont even go to half my line speed
 
The second hop of a tracert shows not the exchange, but rather the place just before getting onto the ISP's network.

Therefore - we can use it to calculate whether the problem lies with the ISP, or whether it lies with the line/exchange

I would argue foul on this. You said it yourself, "but rather the place just before getting onto the ISP's network." It would thus NOT ONLY be the line/exchange as you put it, but it DOES rightfully also include things like IPC, and IPC congestion.

This also is why allot of people on Afrihost has issues with 2nd hop latencies (and Afrihost puppets are quick to jump and say 'exchange'), but as soon as they use a different ISP (different IPC), then the 2nd hop latency is gone. All of a sudden it's not a exchange issue.
 
I would argue foul on this. You said it yourself, "but rather the place just before getting onto the ISP's network." It would thus NOT ONLY be the line/exchange as you put it, but it DOES rightfully also include things like IPC, and IPC congestion.

This also is why allot of people on Afrihost has issues with 2nd hop latencies (and Afrihost puppets are quick to jump and say 'exchange'), but as soon as they use a different ISP (different IPC), then the 2nd hop latency is gone. All of a sudden it's not a exchange issue.

Exactly. Starting to think Zertop wants a job at AH.
 
I am sort of stuck as to where to go

I have sent a lengthy ticket to Afrihost a couple of weeks ago the performance of my line is kinda bad (ticket DUU-647-86661) nothing has improved, port resets etc

I have two accounts with afrihost 2meg business uncapped and a 10 gb capped account (I have a 4 meg line so I was hoping to get full line speed on the capped account)

To quote my ticket



Nothing has changed, the tracert is basically the same and so are general pings to services. YouTube buffers like hell

I have gotten port reset after port reset

Line Stats : http://imgur.com/d91Mi5G

What I dont get is I have a high attainable rate but my line is worse than a fish right now.

Yea, something is a bit fishy there :P Line stats look good, but pings are off :/

Have you tried reporting a fault and getting a technician over?

Z
 
I would argue foul on this. You said it yourself, "but rather the place just before getting onto the ISP's network." It would thus NOT ONLY be the line/exchange as you put it, but it DOES rightfully also include things like IPC, and IPC congestion.

This also is why allot of people on Afrihost has issues with 2nd hop latencies (and Afrihost puppets are quick to jump and say 'exchange'), but as soon as they use a different ISP (different IPC), then the 2nd hop latency is gone. All of a sudden it's not a exchange issue.

I was personally under the impression that the IPC is much further along the trail. However, as I said, Im no expert. However, if there is an IPC issue, the second hop should still be perfect!

My reasoning was this for i.got~issues; he was showing 1000ms pings to both his second hops (Namely the MWEB and Afrihost one's). This proves that the issue ran over both the accounts. Thats why I mentioned a line fault :P

Im quite sure, however, that the second hop excludes the ISP, and is the connection between the exchange and the user. (well, I think i just contradicted what I said previously in the past sentence, but oh well).

While the exchange doesnt show up as a physical IP, it shows up in the form of the first ISP router. So when your line enters the exchange, it gets routed to the ISP it belongs to... if that makes sense.
 
Last edited:
You said Business Uncapped had no shaping. Your Client-Zone says "No Shaping Applied".
The fact that yesterday I could only download half of what I did on Friday, along with the fact that I'm currently downloading at 12kb/s on a 1mb Afrihost Business uncapped account determined that was a lie.

I've been a loyal Afrihost supporter, convincing all my friends on colleagues to go with Afrihost. In the last 6 months, I've tried 1mb uncapped, 1mb business uncapped, 4mb uncapped, 2mb uncapped and now 1mb business uncapped again. I've been dissappointed with every single one of those packages.

I do download substantially, but 99% of all that downloading is from Afrihost's own news server. I never download during peak times. And on top of it all, I call BS on Afrihost throttling rules. 25% throttling on a 4mb line does not in any way equal 12kbps downloads.

P.S. "Real-Time services" like web browsing and youtube are most definitely affected, at the moment. For me anyway.

Hi Elbad

Our Business accounts are definitely not shaped.
Please PM me your account info and any tests you have available so that I can have a look at this and diagnose further.
 
I was personally under the impression that the IPC is much further along the trail. However, as I said, Im no expert. However, if there is an IPC issue, the second hop should still be perfect!


ADSL (your line) -> DSLAM (Telkom) -> BRAS (Telkom) -> ISP (IPC) -> Breakout (ISP's network, peering, transit, etc). -> The Internet

That's how I've been led to understand it. I have seen some very detailed documentation on this though - I just can't remember where. Will have to see if I can dig it out again somewhere.
 
The little grey box is called a DSLAM. Its pretty much the link between the Exchange and the User. The exchange is a physical building in a town where all the telkom lines come into. The "exchange" per se isn't shown, but rather is the middle link. The second hop of a tracert shows not the exchange, but rather the place just before getting onto the ISP's network.

Therefore - we can use it to calculate whether the problem lies with the ISP, or whether it lies with the line/exchange. Each ISP has a different "handler", which accounts for the different IP's :)

I could be wrong, because my knowledge on this isn't the greatest - but someone else should be able to tell you if that is correct :)

If this is the case, then the fact that Afrihost and MWEB are using different buildings, may not meant here's a problem with his line per say. If his line connects to MWEB's exchange fine, but not to Afrihosts, then does that not mean the problem is not with his line, but somewhere between the DSLAM and the Exchange?

In this case, there does seem to be an issue with his line, but it seems worse connecting to the Afrihost Exchange? If Telkom are called out, and the ping to the Telkom Exchange is fine, then who does the fault lie with?
 
ADSL (your line) -> DSLAM (Telkom) -> BRAS (Telkom) -> ISP (IPC) -> Breakout (ISP's network, peering, transit, etc). -> The Internet

That's how I've been led to understand it. I have seen some very detailed documentation on this though - I just can't remember where. Will have to see if I can dig it out again somewhere.

Yea, thats the impression I had.

-> Router (Hop 1 in most cases)
-> ADSL (your line)
-> DSLAM (Telkom)
-> BRAS (Telkom) (Hop 2)
-> ISP (IPC) (Hop 3,4)
-> Breakout (ISP's network, peering, transit, etc). (Hop 5)
-> The Internet (Hop 6)
 
If this is the case, then the fact that Afrihost and MWEB are using different buildings, may not meant here's a problem with his line per say. If his line connects to MWEB's exchange fine, but not to Afrihosts, then does that not mean the problem is not with his line, but somewhere between the DSLAM and the Exchange?

In this case, there does seem to be an issue with his line, but it seems worse connecting to the Afrihost Exchange? If Telkom are called out, and the ping to the Telkom Exchange is fine, then who does the fault lie with?

The exchange is one physical telkom building, and that is where the lines come in to be routed to the different ISP's. That's why when the exchange is congested, these route switches cannot take place fast enough, causing higher latencies.

Thats why It is impossible to tell whether the issue is with the line or with the exchange in one tracert :)
 
Status
Not open for further replies.
Top
Sign up to the MyBroadband newsletter
X