Afrihost Uncapped ADSL Feedback (Pt2)

[Voxratio]

It doesn't say Im getting shaped, but I swear I'm getting roughly 60% of my 4mb line.


EDIT: Just did another speed test. Back to 4mb

 

[Rade84]

I just wish I could stream 480p on my 4mb line. Thats all. Is it so much to ask
Doubt the network is congested at 3am. Still wont stream twitch above 100kBps... good times.

 

[Orihalcon]

I just wish I could stream 480p on my 4mb line. Thats all. Is it so much to ask
Doubt the network is congested at 3am. Still wont stream twitch above 100kBps... good times.

This is a Twitch issue. Google it. Well documented on NUMEROUS sites. A Reddit user posted a guide up on how to fix the problem by running the streams via VLC and that sorted the problem out for me.

 

[AfriMan]

Why is AH shaping Steam downloads now? Last time I needed to get game files, steam was at least flying irrespective of other services, now it's:
Shaping Policy: 18.8% Download/ 6.25% P2P
Steam shows 68.6KB/S CURRENT, it takes more than a day to grab a game's full files...

The other question I have is, AH sent me an email saying congrats, due to new telkom pricing I can upgrade for less or stay on the same line speed : "Uncapped 4096 kbps; DSL (up to 10mbps) Current Sync Rate: 4096kbps" and save R100

I elect to stay on the 4meg line only to be told that I will then also have to stay on the price I am currently paying for the service... What happened to the "congrats you will save" story?

On the first question, most business are back in full swing as of yesterday, so we've seen a fairly big upsurge in usage. We're trying to be pro-active to make sure that we don't see spikes in latency and dips in performance, and we'll continue to adjust and tweak as things settle for the best performance. Remember again that we're mainly talking about the North (Gauteng) region. We're seeing little to no shaping for the moment in the East (Durban) and South (Cape) regions. With regard to Steam, any non-realtime download should be shaped where your shaping policy is active. If Steam has not been shaped in the past when shaping is active on your account, it may have been a mis-diagnosis of the packet signature.

On the second issue, we'll need to see exactly what is transpiring. Remember that your product is the determining factor here. In order to realise a saving, we need to move you from the 10Mbps product (previously Fastest DSL now DSL 3) to the new 4Mbps product (DSL 2) to realise the saving for you. Please drop me a PM so I can check if that was the issue here. Sorry for the confusion though - I promise we're doing our best to decode and demystify this process from our side

 

[AfriMan]

PS: The shaping has been horrendous today. Been stuck on 18% Download / 6% P2P for the entire duration of the day. Network being flooded again?

Most people returned to work yesterday, and business are back in full swing for the moment. As I mentioned on another post, we've being very cautious to make sure that we run at the best performance level, especially for unshaped realtime services, so we're avoiding latency on the network. From what we're seeing, realtime performance should be excellent, but you'll see more shaping that we have over the past 3-4 weeks for now.

Remember that we only received a partial delivery on our IPC upgrade. We hope to have the balance delivered as quickly as possible to have more leeway to shape less in the North especially. There's very little shaping happening in the South & East (if any)

EDIT: Apologies, I can see that we are shaping in South and East over yesterday's reports. Will review the reports more closely in the coming week to give more feedback on how each region is performing.

 

[AfriMan]

Still shaped to 6%? Very strange and sudden @Afriguy? Are you sure the shaper doesn't need a kick into line

As far as we can tell the shaper is working as it should. But we can always take a closer look on a case by case basis.
Drop me a PM and let's have a look at the times in question

 

[AfriMan]

It doesn't say Im getting shaped, but I swear I'm getting roughly 60% of my 4mb line.


EDIT: Just did another speed test. Back to 4mb

Glad it's looking better.

I'm not sure which tests you ran to determine the 60% figure (speedtest or file download). If you see an issue like this give me a little more detail (you can drop me a PM) and I'll be happy to check it out for you

 

[Zertop]

AFRIMAN!!! WELCOME BACK

Hope you holiday was great

 

[AfriMan]

I just wish I could stream 480p on my 4mb line. Thats all. Is it so much to ask
Doubt the network is congested at 3am. Still wont stream twitch above 100kBps... good times.

There are some issues with Twitch that are known. From our own tests and comms with Twitch's service providers, we have optimised from our side as best as possible

 

[AfriMan]

AFRIMAN!!! WELCOME BACK

Hope you holiday was great

Hey, thanks dude. Had an awesome one

Happy New Year to you (and all our MyBB Afripeeps)

 

[Rade84]

This is a Twitch issue. Google it. Well documented on NUMEROUS sites. A Reddit user posted a guide up on how to fix the problem by running the streams via VLC and that sorted the problem out for me.

Not like I havent done some digging on the side... I know twitch has issues. I also know my stream was not nearly this laggy 6 months ago.
The VLC method is cool and all but is having the opposite effect for me. The stream is way more laggy through VLC and wont peak above 70kbps while at the same time the web browser will give me 100kbps (woohoo). Maybe it will improve tonight. P.S. Full speeds on speedtests.

 

[Voxratio]

Glad it's looking better.

I'm not sure which tests you ran to determine the 60% figure (speedtest or file download). If you see an issue like this give me a little more detail (you can drop me a PM) and I'll be happy to check it out for you

I ran the speedtest on the MTN network.

~14ms ping all three tests. The first test it was going just over 2Mbps, I redid it and still got ~2Mbps. 10 minutes later (between time of post and edit), the third test, it went to ~3.5Mbps (which is standard for 4Mb line) .

Haven't seen any shaping applied in the Clientzone panel, even during business hours

 

[AfriMan]

I ran the speedtest on the MTN network.

~14ms ping all three tests. The first test it was going just over 2Mbps, I redid it and still got ~2Mbps. 10 minutes later (between time of post and edit), the third test, it went to ~3.5Mbps (which is standard for 4Mb line) .

Haven't seen any shaping applied in the Clientzone panel, even during business hours

Shaping should not affect a realtime service like a speed test. That would be either contention on the network (but then one would also see high latency), congestion on the exchange, a cable fault on the line or contention on your local network (downloads or services running in the background or on other local PC's or devices).

Trace routes also give a good indication, but also best to make sure that your local network is running optimally and not influencing the tests.

 

[Segg]

Anyone else having high latency tonight?

Microsoft Windows [Version 6.1.7601]
Copyright (c) 2009 Microsoft Corporation. All rights reserved.

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

Tracing route to www.google.co.za [173.194.34.152]
over a maximum of 30 hops:

1 1 ms <1 ms <1 ms 10.0.0.1
2 * 702 ms 148 ms 105-236-6-129-esr-lo.mtnbusiness.co.za [105.236.
6.129]
3 291 ms 479 ms 502 ms ipc-recieve-jh-4a.za.mtnbusiness.net [41.181.178
.41]
4 1098 ms 961 ms 1048 ms qux-jh-dca-2.za-b.za.mtnbusiness.net [41.181.165
.115]
5 1067 ms 1074 ms 1087 ms jh-dca-2.za--qux-b.za.mtnbusiness.net [41.181.16
5.114]
6 1182 ms 1275 ms 1175 ms am-tpr-1.nl--am-cr-1.nl-a.mtn.net [209.212.111.1
41]
7 1349 ms 1355 ms * core2.ams.net.google.com [195.69.145.100]
8 1312 ms * 1165 ms 209.85.254.95
9 425 ms 594 ms 707 ms 72.14.238.69
10 1061 ms * 1260 ms 209.85.240.29
11 1175 ms 1083 ms 1039 ms 209.85.253.197
12 * * * Request timed out.
13 782 ms 797 ms * lhr14s21-in-f24.1e100.net [173.194.34.152]
14 * * 1039 ms lhr14s21-in-f24.1e100.net [173.194.34.152]

Trace complete.

 

[Zertop]

Anyone else having high latency tonight?

Microsoft Windows [Version 6.1.7601]
Copyright (c) 2009 Microsoft Corporation. All rights reserved.

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

Tracing route to www.google.co.za [173.194.34.152]
over a maximum of 30 hops:

1 1 ms <1 ms <1 ms 10.0.0.1
2 * 702 ms 148 ms 105-236-6-129-esr-lo.mtnbusiness.co.za [105.236.
6.129]
3 291 ms 479 ms 502 ms ipc-recieve-jh-4a.za.mtnbusiness.net [41.181.178
.41]
4 1098 ms 961 ms 1048 ms qux-jh-dca-2.za-b.za.mtnbusiness.net [41.181.165
.115]
5 1067 ms 1074 ms 1087 ms jh-dca-2.za--qux-b.za.mtnbusiness.net [41.181.16
5.114]
6 1182 ms 1275 ms 1175 ms am-tpr-1.nl--am-cr-1.nl-a.mtn.net [209.212.111.1
41]
7 1349 ms 1355 ms * core2.ams.net.google.com [195.69.145.100]
8 1312 ms * 1165 ms 209.85.254.95
9 425 ms 594 ms 707 ms 72.14.238.69
10 1061 ms * 1260 ms 209.85.240.29
11 1175 ms 1083 ms 1039 ms 209.85.253.197
12 * * * Request timed out.
13 782 ms 797 ms * lhr14s21-in-f24.1e100.net [173.194.34.152]
14 * * 1039 ms lhr14s21-in-f24.1e100.net [173.194.34.152]

Trace complete.

You have some seeeriiioouuusllyyy bad exchange issues there, your second hop should be under 15ms

 

[Segg]

You have some seeeriiioouuusllyyy bad exchange issues there, your second hop should be under 15ms

Strange that, never ever had an exchange problem before, first time its been that high, I'm hoping a port reset will work

EDIT: Didn't work and now my sync speed is down to 3megs :wtf:

 

[Zertop]

Strange that, never ever had an exchange problem before, first time its been that high, I'm hoping a port reset will work

EDIT: Didn't work and now my sync speed is down to 3megs :wtf:

View attachment 91812

Check that your phone filter is working properly Try removing the phone and seeing what happens

 

[AfriMan]

Strange that, never ever had an exchange problem before, first time its been that high, I'm hoping a port reset will work

EDIT: Didn't work and now my sync speed is down to 3megs :wtf:

View attachment 91812

If you did the reset using Fix My Line - just run it again and it will set your sync rate back to max

 

[AfriMan]

Anyone else having high latency tonight?

Microsoft Windows [Version 6.1.7601]
Copyright (c) 2009 Microsoft Corporation. All rights reserved.

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

Tracing route to www.google.co.za [173.194.34.152]
over a maximum of 30 hops:

1 1 ms <1 ms <1 ms 10.0.0.1
2 * 702 ms 148 ms 105-236-6-129-esr-lo.mtnbusiness.co.za [105.236.
6.129]
3 291 ms 479 ms 502 ms ipc-recieve-jh-4a.za.mtnbusiness.net [41.181.178
.41]
4 1098 ms 961 ms 1048 ms qux-jh-dca-2.za-b.za.mtnbusiness.net [41.181.165
.115]
5 1067 ms 1074 ms 1087 ms jh-dca-2.za--qux-b.za.mtnbusiness.net [41.181.16
5.114]
6 1182 ms 1275 ms 1175 ms am-tpr-1.nl--am-cr-1.nl-a.mtn.net [209.212.111.1
41]
7 1349 ms 1355 ms * core2.ams.net.google.com [195.69.145.100]
8 1312 ms * 1165 ms 209.85.254.95
9 425 ms 594 ms 707 ms 72.14.238.69
10 1061 ms * 1260 ms 209.85.240.29
11 1175 ms 1083 ms 1039 ms 209.85.253.197
12 * * * Request timed out.
13 782 ms 797 ms * lhr14s21-in-f24.1e100.net [173.194.34.152]
14 * * 1039 ms lhr14s21-in-f24.1e100.net [173.194.34.152]

Trace complete.

Latencies are rock solid on all my monitoring and my own connection. Did these improve after the port reset?

 

[Segg]

All working now

Microsoft Windows [Version 6.1.7601]
Copyright (c) 2009 Microsoft Corporation. All rights reserved.

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

Tracing route to www.google.co.za [173.194.34.191]
over a maximum of 30 hops:

1 2 ms 1 ms 1 ms 10.0.0.1
2 33 ms 8 ms 8 ms 105-236-6-129-esr-lo.mtnbusiness.co.za [105.236.
6.129]
3 10 ms 9 ms 10 ms ipc-recieve-jh-1a.za.mtnbusiness.net [41.181.178
.5]
4 10 ms 15 ms 10 ms qux-jh-dca-2.za-b.za.mtnbusiness.net [41.181.165
.115]
5 10 ms 9 ms 9 ms jh-dca-2.za--qux-b.za.mtnbusiness.net [41.181.16
5.114]
6 199 ms 199 ms 198 ms am-tpr-1.nl--am-cr-1.nl-a.mtn.net [209.212.111.1
41]
7 202 ms 203 ms 203 ms core2.ams.net.google.com [195.69.145.100]
8 216 ms 216 ms 216 ms 209.85.254.90
9 204 ms 217 ms 202 ms 209.85.253.249
10 228 ms 199 ms 199 ms 209.85.243.32
11 199 ms 199 ms 199 ms 209.85.244.241
12 216 ms 213 ms 216 ms 209.85.253.175
13 201 ms 201 ms 202 ms lhr14s22-in-f31.1e100.net [173.194.34.191]

Trace complete.



Afriman, any idea what happened? Didn't do anything on my side except reset the port again (23:15) after I checked if the latency had reduced, Ive never had exchange congestion before and I doubt that an exchange can become congested over night

The weird thing from this luckily brief experience is that although my SNR stayed constant, my sync speed dropped, then climbed again As if Telkom are messing around in the area

EDIT: Forgot to mention, the latency gradually reduced to about 700 then 500 then suddenly dropped to what you see