Afrihost Uncapped DSL Feedback

[Reaper1242]

Do know with which Fibre infrastructure provider he is?

Vumatel

But you're still not getting it. If he is seeing the same issue at the same times as we are then it's related to the same problem.
Say what you want about "we need to test" but it's very simple.

Same problem, same time of day, same cause.

 

[bradlesliect]

I'll need to follow-up on this, could you please PM me your details. I wasn't aware of any issues with regards to latency that go back a full month. There was an issue that was resolved last night that was affecting the classification software on our system software that was affecting our Uncapped Clients.

We are also in the process of upgrading our IPC capacity to ensure that our Clients continue to have the best experience possible.

Why must I PM you? You have pages and pages and pages of posts right here! Unfortunately, the agent I spoke to on the phone earlier didnt know he was speaking to a network engineer until I gave him info only a network engineer would know! Your NOC should have picked this issue up and dealt with it when it was found there is a trend being reported! ANSWERS!!!!

 

[SickTricky]

Why must I PM you? You have pages and pages and pages of posts right here! Unfortunately, the agent I spoke to on the phone earlier didnt know he was speaking to a network engineer until I gave him info only a network engineer would know! Your NOC should have picked this issue up and dealt with it when it was found there is a trend being reported! ANSWERS!!!!

I dont think its only afrihost... telkom seems to be unfased for some reason...

 

[bradlesliect]

Vumatel

But you're still not getting it. If he is seeing the same issue at the same times as we are then it's related to the same problem.
Say what you want about "we need to test" but it's very simple.

Same problem, same time of day, same cause.

At the risk of sounding condescending and arrogant...this is not a INFRASTRUCTURE ISSUE....all roads lead to Afrihost which points to routing, traffic priority problems. Its simple. Log change control and UNDO the changes you made on the network thats affecting ALL your clients.

You've stuffed up your routing policies and prioritizing the incorrect traffic type at the wrong time of day which is causing congestion, putting strain on your network and its just a matter of time before your edge/gateway routers fall over because it can't handle the influx of routing requests and then its just a matter of time before some of your IP Ranges are blocked by upstreams due to flooding. FIX! FIX! FIX!

 

[Reaper1242]

My guess is that there's a big problem and they don't know how to fix it. All of this is just to buy time and create the false impression that they're fixing it.

Now that people are figuring it out it's not going to be so easy.

 

[nerevar]

Gotta just go on what most here is saying, cant remember the last time I have had issue with Afrihost, and the last 10 days or so its been a roller coaster

 

[bradlesliect]

My guess is that there's a big problem and they don't know how to fix it. All of this is just to buy time and create the false impression that they're fixing it.

Now that people are figuring it out it's not going to be so easy.

....and they have been stringing clients along for more that month. By the looks of things each client has been treated as an "isolated case" whilst back at the ranch, they are applying #quick_fix to try and resolve and its probably been a case of "hit and miss".

AH - Call in the big guys. The dudes you pay them BIG dollars too and let them fix and stabilize your network again. Let them work for their money...

tick, tock...the clock is ticking...

 

[AfriGenie]

Vumatel

But you're still not getting it. If he is seeing the same issue at the same times as we are then it's related to the same problem.
Say what you want about "we need to test" but it's very simple.

Same problem, same time of day, same cause.

That's not the case though, there is significant enough differences between a DSL product with us and a Fibre product on Vumatel that it's not likely to be the same issue.

To be sure though we'll need to check on this actual account to see what's going.

 

[Reaper1242]

That's not the case though, there is significant enough differences between a DSL product with us and a Fibre product on Vumatel that it's not likely to be the same issue.

To be sure though we'll need to check on this actual account to see what's going.

Excuse me if I'm dubious of anything coming from AH right now.

 

[MoonOwl]

Tonight, it's working as expected for me. An improvement from last night and the times before. Thank you, Afrihost, for rectifying the issue.

However, I do wish the broadband status shown on your website was indicative of the reality on the ground. While the spike in demand on your network was existent, it would have been more faithful to show the reality of the network.

I trust you will dig yourselves out of the hole you currently are in

 

[happydaddy]

When i tracert to afrihost.co.za it goes to Mweb Why is this? Pings to afrihost is good but outside of afrihost other local site the ping is bad.

Microsoft Windows [Version 6.1.7600]
Copyright (c) 2009 Microsoft Corporation. All rights reserved.

C:\Users\happydaddy>tracert www.afrihost.co.za

Tracing route to www.afrihost.co.za [41.86.112.68]
over a maximum of 30 hops:

1 8 ms <1 ms <1 ms 192.168.20.1
2 2 ms 2 ms 2 ms 10.0.4.2
3 5 ms 3 ms 4 ms 10.0.56.1
4 5 ms 6 ms 5 ms 10.0.48.1
5 * * * Request timed out.
6 61 ms 31 ms 28 ms cpt-up1.ip.adsl.co.za [169.1.5.105]
7 82 ms 72 ms 61 ms cpt-in1.ip.adsl.co.za [169.1.5.118]
8 136 ms 124 ms 131 ms cpt-net1.ip.adsl.co.za [169.1.5.128]
9 152 ms 135 ms 109 ms optinet.ixp.capetown [196.10.140.15]
10 56 ms 66 ms 58 ms te0-0-0-1.cpt-p-1.optinet.net [197.84.7.1]
11 57 ms 69 ms 49 ms te0-7-0-3.cpt-p-2.optinet.net [197.84.7.2]
12 49 ms 55 ms 48 ms gi0-20.trv-pe-2.optinet.net [197.84.4.17]
13 67 ms 64 ms 64 ms 197-84-4-101.cpt.mweb.co.za [197.84.4.101]
14 49 ms 54 ms 55 ms 197-80-7-15.jhb.mweb.co.za [197.80.7.15]
15 50 ms 44 ms 57 ms 197-81-226-81.jhb.mweb.co.za [197.81.226.81]
16 56 ms 49 ms 56 ms 197-81-226-61.jhb.mweb.co.za [197.81.226.61]
17 61 ms 66 ms 50 ms 197-81-229-4.jhb.mweb.co.za [197.81.229.4]
18 53 ms 49 ms 48 ms 197-81-229-11.jhb.mweb.co.za [197.81.229.11]
19 53 ms 44 ms 44 ms 41-86-112-68.mweb.co.za [41.86.112.68]

Trace complete.

C:\Users\happydaddy>tracert www.afrihost.co.za

 

[AfriGenie]

....and they have been stringing clients along for more that month. By the looks of things each client has been treated as an "isolated case" whilst back at the ranch, they are applying #quick_fix to try and resolve and its probably been a case of "hit and miss".

AH - Call in the big guys. The dudes you pay them BIG dollars too and let them fix and stabilize your network again. Let them work for their money...

tick, tock...the clock is ticking...

As I said earlier, we are in the process of rolling out additional IPC capacity which make sure that we have enough bandwidth for the demand that we are currently seeing. Our Network Team identified this last week and we ordered the additional capacity to have it go live as quickly as possible.

The classification issue was identified in the last two days and our Team implemented the fixes last night.

If your performance is still not great, then I'll need to check on your account itself to see what's going on.

 

[AfriGenie]

Tonight, it's working as expected for me. An improvement from last night and the times before. Thank you, Afrihost, for rectifying the issue

Thanks for the update.

 

[AfriGenie]

When i tracert to afrihost.co.za it goes to Mweb Why is this? Pings to afrihost is good but outside of afrihost other local site the ping is bad.

Afrihost.com makes use of Cloudflare, and there is a Cloudflare CDN hosted in the Mweb Data Center.

To which site or service do you see latency at the moment?

 

[bradlesliect]

As I said earlier, we are in the process of rolling out additional IPC capacity which make sure that we have enough bandwidth for the demand that we are currently seeing. Our Network Team identified this last week and we ordered the additional capacity to have it go live as quickly as possible.

The classification issue was identified in the last two days and our Team implemented the fixes last night.

If your performance is still not great, then I'll need to check on your account itself to see what's going on.

FINALLY!!!! SOME TRUTH!!!!

Why not just admit right from the beginning that your network is packed to capacity and you using your clients as guinea pigs to see what works and what doesn't?!!! Its clear that routing policies that have been applied to alleviate bandwidth issues have had the opposite effect and this is all because its probably not been THOROUGHLY tested in a lab environment taking all the variables into account?

#this is just how I see it right now with the little bit of knowledge I have.

Again...how are you going to fix CURRENT situation until permanent solution is found? Take a closer look at your routing policies, access-lists, traffic priorities? For a NOC such as yours, this should be a walk in the park?

 

[bradlesliect]

Afrihost.com makes use of Cloudflare, and there is a Cloudflare CDN hosted in the Mweb Data Center.

To which site or service do you see latency at the moment?

Dude...your network is a mess. PERIOD! Latency everywhere. Afrihost should be paying their clients for being their clients!

 

[Soprono]

FINALLY!!!! SOME TRUTH!!!!

Why not just admit right from the beginning that your network is packed to capacity and you using your clients as guinea pigs to see what works and what doesn't?!!! Its clear that routing policies that have been applied to alleviate bandwidth issues have had the opposite effect and this is all because its probably not been THOROUGHLY tested in a lab environment taking all the variables into account?

#this is just how I see it right now with the little bit of knowledge I have.

Again...how are you going to fix CURRENT situation until permanent solution is found? Take a closer look at your routing policies, access-lists, traffic priorities? For a NOC such as yours, this should be a walk in the park?

They admitted this yesterday already with some clarification this morning of a correction plan.

Guess after admitting it and getting people to work together they could better pinpoint some issues quicker.

 

[happydaddy]

Below 2 sites i tracert to

Microsoft Windows [Version 6.1.7600]
Copyright (c) 2009 Microsoft Corporation. All rights reserved.

C:\Users\happydaddy>tracert www.google.co.za

Tracing route to www.google.co.za [216.58.223.35]
over a maximum of 30 hops:

1 8 ms <1 ms <1 ms 192.168.20.1
2 2 ms 1 ms 2 ms 10.0.4.2
3 4 ms 4 ms 5 ms 10.0.56.1
4 5 ms 5 ms 5 ms 10.0.48.1
5 * * * Request timed out.
6 31 ms 26 ms 35 ms cpt-up1.ip.adsl.co.za [169.1.5.105]
7 56 ms 64 ms 56 ms 169-1-21-66.ip.afrihost.co.za [169.1.21.66]
8 127 ms 129 ms 146 ms cpt-net1.ip.adsl.co.za [169.1.5.128]
9 51 ms 42 ms 51 ms 169-1-21-29.ip.afrihost.co.za [169.1.21.29]
10 265 ms 94 ms 97 ms 169-1-5-243.ip.afrihost.co.za [169.1.5.243]
11 55 ms 69 ms 62 ms 72.14.239.129
12 176 ms 184 ms 163 ms jnb01s08-in-f35.1e100.net [216.58.223.35]

Trace complete.

C:\Users\happydaddy>

Microsoft Windows [Version 6.1.7600]
Copyright (c) 2009 Microsoft Corporation. All rights reserved.

C:\Users\happydaddy>tracert www.google.co.za

Tracing route to www.google.co.za [216.58.223.35]
over a maximum of 30 hops:

1 8 ms <1 ms <1 ms 192.168.20.1
2 2 ms 1 ms 2 ms 10.0.4.2
3 4 ms 4 ms 5 ms 10.0.56.1
4 5 ms 5 ms 5 ms 10.0.48.1
5 * * * Request timed out.
6 31 ms 26 ms 35 ms cpt-up1.ip.adsl.co.za [169.1.5.105]
7 56 ms 64 ms 56 ms 169-1-21-66.ip.afrihost.co.za [169.1.21.66]
8 127 ms 129 ms 146 ms cpt-net1.ip.adsl.co.za [169.1.5.128]
9 51 ms 42 ms 51 ms 169-1-21-29.ip.afrihost.co.za [169.1.21.29]
10 265 ms 94 ms 97 ms 169-1-5-243.ip.afrihost.co.za [169.1.5.243]
11 55 ms 69 ms 62 ms 72.14.239.129
12 176 ms 184 ms 163 ms jnb01s08-in-f35.1e100.net [216.58.223.35]

Trace complete.

C:\Users\happydaddy>

 

[bradlesliect]

They admitted this yesterday already with some clarification this morning of a correction plan.

Guess after admitting it and getting people to work together they could better pinpoint some issues quicker.

Yesterday? Where? There is NOTHING on their notices page. This would have been an admission of a problem being experienced. No email comms sent to clients, no SMS, no forum posts.

Anyways...I've made my point. I shall now be quiet and give them time till morning to come up with a PROPER comms to ALL their clients, explaining current situation and what is being done to remedy it. Hopefully its not too little...too late...

 

[bradlesliect]

If your performance is still not great, then I'll need to check on your account itself to see what's going on.

My account has already been checked, checked and re-checked. There are no restrictions, throttling or anything that would have been implemented / applied because I contravened AH network AUP.