Afrihost Capped DSL Feedback

roadcat · Mar 14, 2017

steel_dog said:
Don't forget that it's not just my connection. There are a lot of other folks reporting exactly the same. I've been diagnosing in the best interest of all.

You articulate the issue so nicely, I'll follow

AfriGenie · Mar 14, 2017

steel_dog said:
Don't forget that it's not just my connection. There are a lot of other folks reporting exactly the same. I've been diagnosing in the best interest of all.

Once the Critical Care Team has figured out what's going on with your connection specifically, they'll be able to apply the fix to other affected Clients.

AfriGenie · Mar 14, 2017

homerj said:
Yes

Wow... That's strange... Are you having problems with any other News Servers?

steel_dog · Mar 14, 2017

AfriGenie said:
Wow... That's strange... Are you having problems with any other News Servers?

Checkout my post about the news server. It's #915.

roadcat · Mar 14, 2017

Oh dear..

Normal = 30 down.

Dodgy SAIX server

Mweb Server

BRAS
|------------------------------------------------------------------------------------------|
| WinMTR statistics |
| Host - % | Sent | Recv | Best | Avrg | Wrst | Last |
|------------------------------------------------|------|------|------|------|------|------|
| 192.168.1.1 - 0 | 411 | 411 | 0 | 0 | 2 | 0 |
| 155.239.255.250 - 0 | 411 | 411 | 16 | 21 | 36 | 19 |
|________________________________________________|______|______|______

AfriGenie · Mar 14, 2017

roadcat said:
Oh dear..

Normal = 30 down.

Dodgy SAIX server
View attachment 423634

Mweb Server
View attachment 423636

BRAS
|------------------------------------------------------------------------------------------|
| WinMTR statistics |
| Host - % | Sent | Recv | Best | Avrg | Wrst | Last |
|------------------------------------------------|------|------|------|------|------|------|
| 192.168.1.1 - 0 | 411 | 411 | 0 | 0 | 2 | 0 |
| 155.239.255.250 - 0 | 411 | 411 | 16 | 21 | 36 | 19 |
|________________________________________________|______|______|______

That's quite a difference... And you made no changes between those two speedtests?

roadcat · Mar 14, 2017

AfriGenie said:
That's quite a difference... And you made no changes between those two speedtests?

No changes, roughly 5 seconds from one ending to the other starting, the SAIX one could be peering between AH and SAIX? Of more concern is my latency to Cape Town on both speedtests.

EDIT: I am on con-call with the US at the moment and the quality is fine, so I am a little confused. Looking at some traceroutes to different locations, they all seem to take different routes through Cape Town, so maybe webex skips that bad hop.

Hummercellc · Mar 14, 2017

Local latency in the mid hundreds again tonight. East IPC is taking strain again. Sigh....

Geoff.D · Mar 14, 2017

I notice that internet access generally has just about ground to a halt. The problem appears to have started at about 19:45.

That is the same time that a very large win 10 update download kicked in.
Coincidence?

Johnatan56 · Mar 14, 2017

roadcat said:
Dodgy SAIX server
View attachment 423634

Is that the Durban one? Have the same issue on mobile data and other providers.
(Sorry, just posting this to help roadcat)

roadcat · Mar 14, 2017

Johnatan56 said:
Is that the Durban one? Have the same issue on mobile data and other providers.
(Sorry, just posting this to help roadcat)

No, mine are PE to Cape Town tests

Johnatan56 · Mar 14, 2017

roadcat said:
No, mine are PE to Cape Town tests

Streaming a music video at the same time though:

AfriFella · Mar 15, 2017

Good morning guys. Thank you for all your tests last night, we passed a lot of info onto our NOC Team and I believe that they made a few small configuration changes in the South. We'll be monitoring closely and will appreciate your further feedback going forward.

Kensei2605 · Mar 15, 2017

steel_dog said:

Okay here's the tracert via my server ...
As you can see it's going to the Netherlands server.
Forgot the packet loss for now. That's another issue we haven't quite had answers to yet.

Code:

                                       Packets               Pings
 Host                                Loss%   Snt   Last   Avg  Best  Wrst StDev
 1. router.asus.com                   0.0%    41    0.3   0.3   0.2   0.6   0.0
 2. ???
 3. cpt-up1.ip.adsl.co.za             4.9%    41   20.8  23.8  20.1  83.3  10.4
 4. cpt-in1.ip.adsl.co.za             0.0%    41   22.1  24.5  20.2  76.5  11.2
 5. cpt-net1.ip.adsl.co.za            0.0%    41   27.0  55.1  19.6 178.0  40.7
 6. 41.164.52.40                      2.4%    41   20.3  26.6  19.7  56.0   9.5
 7. ix-ae-7-0.tcore2.KLT-Cape-Town.a  0.0%    41   20.8  30.6  20.1 129.6  23.7
 8. if-ae-2-2.tcore1.KLT-Cape-Town.a  0.0%    41  210.2 228.9 172.5 285.6  25.6
 9. if-ae-5-2.tcore2.PV9-Lisbon.as64  4.9%    41  174.6 179.4 173.9 227.5  13.2
10. if-ae-2-2.tcore1.PV9-Lisbon.as64  2.4%    41  175.7 179.4 174.4 227.3  12.3
11. if-ae-1-3.tcore1.SV8-Highbridge.  0.0%    41  180.9 189.0 180.9 219.9  10.9
12. if-ae-19-2.tcore2.L78-London.as6 12.5%    41  226.8 228.6 175.0 264.6  20.8
13. if-ae-8-2.tcore2.AV2-Amsterdam.a  0.0%    41  181.2 186.0 180.8 236.1  11.3
14. if-ae-2-2.tcore1.AV2-Amsterdam.a  0.0%    41  261.4 191.7 174.1 261.4  19.9
15. 195.219.194.118                   0.0%    40  175.5 178.1 174.6 203.0   7.0
16. 1-1.r2.am.hwng.net                0.0%    40  241.6 239.3 180.5 303.7  25.6
17. 81.171.92.236                     0.0%    40  175.7 178.7 174.8 213.8   7.7

Now look at Windows ...

Code:

Target Name: news.afrihost.com
         IP: 169.1.1.36
  Date/Time: 14 Mar 2017 4:57:52 PM - 14 Mar 2017 5:07:52 PM

Hop  Sent  PL%    Min     Max     Avg  Host Name / [IP]
  1    92    0   0.71    1.53    1.01  router.asus.com [192.168.1.1]
  2    92  100      0       0       0   [-]
  3    92    5  20.39   70.29   22.87  cpt-up1.ip.adsl.co.za [169.1.5.105]
  4    92    0  20.00   65.41   23.33  cpt-up2.ip.adsl.co.za [169.1.5.106]
  5    92    0  19.96  191.16   49.00  cpt-net1.ip.adsl.co.za [169.1.5.128]
  6    92    7  20.21   52.33   22.07  41.164.52.40 [41.164.52.40]
  7    92    0  40.00   62.59   41.86  172.18.1.164 [172.18.1.164]
  8    92    3  41.41  244.91  102.92  41.169.57.219 [41.169.57.219]
  9    92    0  39.17   78.21   42.02  169-1-5-195.ip.afrihost.co.za [169.1.5.195]
 10    92   33  40.56  210.69   60.17  jhb-ex0b.ip.adsl.co.za [169.1.5.63]
 11    92    0  39.84  220.08   54.86  news.afrihost.com [169.1.1.36]

That's going to the local mirror.

Why so? Dunno. Bizzarre.
I flushed my Windows DNS tables and...

Code:

Target Name: news.afrihost.com
         IP: 81.171.92.222
  Date/Time: 14 Mar 2017 5:06:48 PM - 14 Mar 2017 5:16:48 PM

Hop  Sent  PL%     Min     Max     Avg  Host Name / [IP]
  1   172    0    0.70    2.15    0.97  router.asus.com [192.168.1.1]
  2   172  100       0       0       0   [-]
  3   172   11   20.53   65.64   23.82  cpt-up1.ip.adsl.co.za [169.1.5.105]
  4   172    3   20.00   66.48   23.32  169-1-21-66.ip.afrihost.co.za [169.1.21.66]
  5   172    1   20.36  169.93   50.91  cpt-net1.ip.adsl.co.za [169.1.5.128]
  6   172    5   20.36   54.05   22.89  41.164.52.40 [41.164.52.40]
  7   172    0   20.88   72.69   23.71  ix-ae-7-0.tcore1.KLT-Cape-Town.as6453.net [41.206.164.53]
  8   172    0  168.57  211.83  170.93  if-ae-5-2.tcore2.PV9-Lisbon.as6453.net [80.231.159.89]
  9   172    0  168.59  216.32  171.31  if-ae-2-2.tcore1.PV9-Lisbon.as6453.net [80.231.158.5]
 10   172    0  175.22  339.83  192.28  if-ae-1-3.tcore1.SV8-Highbridge.as6453.net [80.231.158.30]
 11   172   84  168.00  267.90  190.69  if-ae-19-2.tcore2.L78-London.as6453.net [80.231.138.22]
 12   172    1  174.75  214.52  177.67  if-ae-8-2.tcore2.AV2-Amsterdam.as6453.net [80.231.131.6]
 13   172    1  167.83  319.29  185.07  if-ae-2-2.tcore1.AV2-Amsterdam.as6453.net [195.219.194.5]
 14   172    3  167.78  222.57  172.04  195.219.194.122 [195.219.194.122]
 15   172    0  169.01  332.69  216.14  1-1.r2.am.hwng.net [69.16.191.101]
 16   172    0  173.00  219.00  176.25  news.afrihost.com [81.171.92.222]

Lo and behold .... it now hits the Nertherlands server(s).

So there are different servers locally and international and I'm guessing the local one ain't working.
If you're hitting that you're likely buggered.

If I'm right about this and the other issues then AH need to start thinking about an engagement with SD.

Additional note: My DNS tables on Windows were flushed 48hrs ago so the local server hit happened with pretty fresh DNS data. Why the server and Windows have different resolutions... not 100% sure yet.

Thanks for all the info will run a trace and reply. Just hate messing with my network setup as everything is working now wifi extenders, android box, ps3 etc change one ip or setting and everything goes bonkers. Will prob then just stay with Sunnyusenet.

Kensei2605 · Mar 15, 2017

Okay did the same test with the same results. Unfortunately flushdns did not change my windows tracert result but funny enough my last hop is fakenews.net.afrihost.co.za.

kombi-man · Mar 15, 2017

AfriGenie said:
Once the Critical Care Team has figured out what's going on with your connection specifically, they'll be able to apply the fix to other affected Clients.

AfriGenie, let me tell you what pi$$es me off about this! There's CLEARLY an issue with AH's capped account, I've been blurting out about this on my own post and on social media, but, five or six of your support engineers respond to the logged calls with comments like, exchange congestion, or telkom line issues, or junk that's "external" to AH's system.

SURELY at some point, AH engineers realize there is indeed an issue, and should either STATE this on their client website, or on the FB page, or somewhere, where dilligent customers then know they don't need to go about re-designing the wheel to find the problem..

In my opinion, I'd have THAT much more respect for the Company for admitting, and publishing the fact there's an issue, instead of telling customers it's their problem. Just my two cents on this.

Last night I tested for the Critical care team, didn't have time to do tracert's etc, but the ookla speedtest at 22:05 on AH was down at sub 2mbps on my 10 MB synced line, and changing to Webafrica and running the same test 3 minutes later, it runs straight at 9.8 Mbps.

The evening before, I did exactly the same test however used an axxess test account the technician asked me to test, and even tho that's on the same Network, it was at 9.7mbps, and the Ah speed was 1.5 mbps!

Here's hoping AH discover the issue very soon. For now I'm using my webafrica account and happydays..

AfriFella · Mar 15, 2017

APaterson said:
AfriGenie, let me tell you what pi$$es me off about this! There's CLEARLY an issue with AH's capped account, I've been blurting out about this on my own post and on social media, but, five or six of your support engineers respond to the logged calls with comments like, exchange congestion, or telkom line issues, or junk that's "external" to AH's system.

SURELY at some point, AH engineers realize there is indeed an issue, and should either STATE this on their client website, or on the FB page, or somewhere, where dilligent customers then know they don't need to go about re-designing the wheel to find the problem..

In my opinion, I'd have THAT much more respect for the Company for admitting, and publishing the fact there's an issue, instead of telling customers it's their problem. Just my two cents on this.

Last night I tested for the Critical care team, didn't have time to do tracert's etc, but the ookla speedtest at 22:05 on AH was down at sub 2mbps on my 10 MB synced line, and changing to Webafrica and running the same test 3 minutes later, it runs straight at 9.8 Mbps.

The evening before, I did exactly the same test however used an axxess test account the technician asked me to test, and even tho that's on the same Network, it was at 9.7mbps, and the Ah speed was 1.5 mbps!

Here's hoping AH discover the issue very soon. For now I'm using my webafrica account and happydays..

Thanks for your feedback. I'm sorry to hear that this is such a frustrating experience.

Our Critical Care Team will always remain by far the best point of contact. Whether it is fixing line related issues or reporting general issues to our Network Team.

As we have discussed before, your issue isn't related to shaping as this will result in poor tests results no matter the server you test to. Our Network Team have made some configuration changes to our network in the South last night and we are monitoring performance going forward. I'd like to encourage you to keep working with our CC Team until a resolution to your problem has been managed.

steel_dog · Mar 15, 2017

Kensei2605 said:
Okay did the same test with the same results. Unfortunately flushdns did not change my windows tracert result but funny enough my last hop is fakenews.net.afrihost.co.za.

Try this. I did last night. Set your server to 81.171.92.222.
In terms of flushing the DNS. You're likely using another DNS server. I've automatically got 169.1.1.2 and 169.1.1.9.
So the issues are twofold.
1. The local news server is not working correctly
2. Why is DNS serving up different results to different users
That's for the AfriPeeps to pass on to their team.

steel_dog · Mar 15, 2017

AfriFella said:
Thanks for your feedback. I'm sorry to hear that this is such a frustrating experience. Our Critical Care Team will always remain by far the best point of contact. Whether it is fixing line related issues or reporting general issues to our Network Team.

As we have discussed before, your issue isn't related to shaping as this will result in poor tests results no matter the server you test to. Our Network Team have made some configuration changes to our network in the South last night and we are monitoring performance going forward. I'd like to encourage you to keep working with our CC Team until a resolution to your problem has been managed.

I absolutely understand APaterson's outlook and frustration. Sounds exactly like I felt back in the pre-POC days when clients were reporting issues only to have AH attempt to answer them away. As it turned out then - the customer was right. Would prefer a bit more circumspection from AH when dealing with customer concerns / observations (especially the customers on this forum).

Anyhow it seems as if the relevant team has been made aware. I guess as a customer I'd like to know what they've found, or suspect, and an ETC, or plan of action, for the resolution.

So we don't lose sight of the issues...
1. NNTP
2. DSTV
3. Slow streaming
and related to that
4. Packet Loss
5. High Latency
Not forgetting - the SAIX test account does NOT exhibit any of the behaviour seen on the AH capped accounts.

roadcat · Mar 15, 2017

I think it's worse

Same to server, using Speedtest

Afrihost to Cape Town

Plugg(IS) to Cape Town

Didn't bother including BRAS trace

Join the MyBroadband community

Get started

Afrihost Capped DSL Feedback

Expert Member

Afrihost

Afrihost

Senior Member

Expert Member

Afrihost

Expert Member

Expert Member

Honorary Master

Honorary Master

Expert Member

Honorary Master

Afrihost

Senior Member

Senior Member

Senior Member

Afrihost

Senior Member

Senior Member

Expert Member