Afrihost Business Uncapped Feedback - 2

Status
Not open for further replies.
I think a more direct approach is needed ....

Afriperson ... Cut the cr@p

The latency is screwy, please fix it. I am quite a few people from all sorts of areas would appreciate it.

We know that there are some problems at the moment, and latency and slow throughput would be the symptoms experienced when these problems cause demand to be less managed than it should be. Once we have this under control, I expect everything will come back to normal. Right now, we're seeing intermittent improvement as we make changes, but I'll definitely report when everything is fully resolved.
 
Well ... dunno

Tracing route to www.afrihost.com [197.242.144.102]
over a maximum of 30 hops:

1 1 ms 1 ms 1 ms 192.168.1.1
2 264 ms 282 ms 408 ms 105-236-1-1-esr-lo.mtnbusiness.co.za [105.236.1.
1]
3 378 ms 275 ms 344 ms 41.181.221.214
4 504 ms 401 ms 415 ms rb-dca-2.za--qux-b.za.mtnbusiness.net [41.181.16
5.118]
5 154 ms 364 ms 275 ms rb-cr-1.za--rb-dca-2.mtnns.net [196.44.31.122]
6 347 ms 390 ms 344 ms 196.44.31.97
7 296 ms 331 ms 233 ms 196.31.220.12
8 179 ms 300 ms 279 ms 196.31.220.25
9 299 ms 326 ms 355 ms 196.31.63.206
10 397 ms 195 ms 110 ms vl9-ah-ha-2.za.mtnbusiness.net [196.30.42.133]
11 315 ms 293 ms 304 ms cms-gm.afrihost.com [197.242.144.102]

Trace complete.

This is not necessarily on our side though. You should only see latency on the 3rd hop. Remember that what you're seeing as the second hop is at the BRAS level, before you reach our network :(
 
Joy of joys, 10M cpt

DrnF1Zn.png


Yet

http://www.speedtest.net/my-result/4230475715

That doesn't look so great. With some of the issues we're seeing with our traffic management, it's very possible that small packets can get through quickly but overall throughput could be poor when demand is high. Until we get our ruleset 100% back on track, this could be the case :(
 
I already posted tests last night. Which other tests do you need?

Guessing I'm not gonna get a reply to this anytime soon.

Saturday was ****ty ****ty ****ty. Yesterday it was fine all day. Guessing your systems function better the colder the weather - sounds absurd, but it's a better guess than anything you've told me so far.
 
May I ask - the "fix my line" on cz is gone my side - thus I am unable to do any port recreates. Is there an alternative besides phoning in to AH?

You can PM one of us to do it. We should restore the feature soon once Telkom have given us an updated API for this and other line management tools :)
 
I think I might be having some issues with congestion recently.

http://pastebin.com/4h3aL2ja

Any way to help get this sorted out?

I can't say this is definitely congestion because we're seeing some increased latency on our network at peak times. We're busy working on this, it appears to be linked to our shaping system and traffic rules.

However, I do see you have a 9ms ping to your router. If this is over a LAN cable it could indicate a problem on the local network like a background service. If it's over WiFi, it's a still a little high but maybe not the cause.
 
Congestion early hours of Sunday morning?

It depends on your area. For some areas, that makes sense for the people in that region. Generally, we see a slightly different usage pattern over the entire network on a weekend but there are still periods of very high demand.
 
Well honesty I have no idea whats going on there. It was happening late on Thursday night as well, even when I switched accounts.'


Edit: Tried another account now and still the same thing...
The weird thing is that its not consistently high, its spiking on the second hop, its up and down all the time.

This is usually an indication of congestion, depending on what your second hop is in your router/network configuration. In most installations, the second hop is the Telkom BRAS that routes traffic to our network.
 
AfriGuy, please could you get someone to look into what is going on?

I am now routing through a Different ESR, and there is a Extra Hop in the trace routes (41.181.221.213)

From Jan 2014 to Dec 2014 this is how my latency and trace routes looked

Traceroute output:
1 105-236-10-1-esr-lo.mtnbusiness.co.za (105.236.10.1) 4.659 ms 4.977 ms 4.639 ms
2 41.181.221.214 (41.181.221.214) 15.572 ms 15.203 ms 16.025 ms
3 rb-dca-2.za--qux-b.za.mtnbusiness.net (41.181.165.118) 15.801 ms 14.793 ms 16.845 ms
4 rb-dca-2.za--rb-cr-1.mtnns.net (196.44.31.122) 17.653 ms 16.728 ms 19.023 ms
5 196.44.31.97 (196.44.31.97) 16.944 ms 15.824 ms 16.030 ms
6 196.31.220.10 (196.31.220.10) 18.387 ms 18.508 ms 15.357 ms
7 196.31.220.21 (196.31.220.21) 16.002 ms 15.546 ms 15.878 ms
8 196.31.63.202 (196.31.63.202) 15.855 ms 15.534 ms 15.714 ms
9 vl9-ah-ha-2.za.mtnbusiness.net (196.30.42.133) 16.017 ms 16.685 ms 15.764 ms
10 cms-gm.afrihost.com (197.242.144.102) 15.572 ms 15.893 ms 15.514 ms

They now look like this, for over a year everything was perfect?

Traceroute output:
1 105-237-134-1-esr-lo.mtnbusiness.co.za (105.237.134.1) 4.977 ms 4.833 ms 4.558 ms
2 41.181.221.213 (41.181.221.213) 62.919 ms 91.475 ms 113.804 ms
3 41.181.221.214 (41.181.221.214) 86.399 ms 43.653 ms 43.348 ms
4 rb-dca-2.za--qux-b.za.mtnbusiness.net (41.181.165.118) 37.348 ms 48.771 ms 40.217 ms
5 rb-cr-1.za--rb-dca-2.mtnns.net (196.44.31.122) 37.944 ms 38.425 ms 37.678 ms
6 196.44.31.97 (196.44.31.97) 43.345 ms 43.061 ms 55.625 ms
7 196.31.220.12 (196.31.220.12) 37.929 ms 61.309 ms 44.202 ms
8 196.31.220.25 (196.31.220.25) 45.622 ms 38.197 ms 37.077 ms
9 196.31.63.206 (196.31.63.206) 49.728 ms 44.480 ms 43.347 ms
10 vl9-ah-ha-2.za.mtnbusiness.net (196.30.42.133) 43.928 ms 42.520 ms 43.652 ms
11 cms-gm.afrihost.com (197.242.144.102) 46.500 ms 44.067 ms 64.012 ms

And before you blame my exchange or line, this is what WebAfrica looks likes

Tracing route to is.co.za [196.26.2.114]
over a maximum of 30 hops:

1 * * * Request timed out.
2 8 ms 8 ms 8 ms 196.38.74.246
3 8 ms 7 ms 7 ms cdsl1-umh-gi0-0-0-3102.ip.isnet.net [196.38.74.245]
4 8 ms 8 ms 8 ms 196.26.210.200
5 8 ms 8 ms 8 ms mi-za-umh-p8-ten-0-0-0-1.ip.isnet.net [168.209.93.240]
6 8 ms 8 ms 8 ms mi-za-umh-p7-ten-0-0-0-1.ip.isnet.net [168.209.93.203]
7 18 ms 19 ms 18 ms mi-za-bry-p7-lo0.ip.isnet.net [168.209.255.208]
8 18 ms ms 18 ms core2a-bry-gi1-0-0.ip.isnet.net [168.209.218.2]
9 18 ms 19 ms 18 ms bry.it.is.co.za [196.14.64.36]

AfriGenie has given no real support what so ever, just generic answers..... I have PM'ed him a number of trace routes and tested with his account which make NO difference what so ever! I have logged faults with Telkom and Charl (TelkomZA) has tested and confirmed this a problem on the MTN network and the routes need to be fixed from MTN's side as all other ISP's work fine.

You posted on Friday but this is where it stopped?? 6262
 
I can't say this is definitely congestion because we're seeing some increased latency on our network at peak times. We're busy working on this, it appears to be linked to our shaping system and traffic rules.

However, I do see you have a 9ms ping to your router. If this is over a LAN cable it could indicate a problem on the local network like a background service. If it's over WiFi, it's a still a little high but maybe not the cause.

I am on a LAN connection, sitting 2m away from my router. Thing is in that table the 9ms was the highest but the average response time was 0 as per the table, so I really don't think the problem is there.


This is usually an indication of congestion, depending on what your second hop is in your router/network configuration. In most installations, the second hop is the Telkom BRAS that routes traffic to our network.


Okay that I can understand, since it occurred on another ISP, but what can be done to resolve this now? You giving me nice explanations but not many solutions...
 
AfriGuy, please could you get someone to look into what is going on?

I am now routing through a Different ESR, and there is a Extra Hop in the trace routes (41.181.221.213)

From Jan 2014 to Dec 2014 this is how my latency and trace routes looked

Traceroute output:
1 105-236-10-1-esr-lo.mtnbusiness.co.za (105.236.10.1) 4.659 ms 4.977 ms 4.639 ms
2 41.181.221.214 (41.181.221.214) 15.572 ms 15.203 ms 16.025 ms
3 rb-dca-2.za--qux-b.za.mtnbusiness.net (41.181.165.118) 15.801 ms 14.793 ms 16.845 ms
4 rb-dca-2.za--rb-cr-1.mtnns.net (196.44.31.122) 17.653 ms 16.728 ms 19.023 ms
5 196.44.31.97 (196.44.31.97) 16.944 ms 15.824 ms 16.030 ms
6 196.31.220.10 (196.31.220.10) 18.387 ms 18.508 ms 15.357 ms
7 196.31.220.21 (196.31.220.21) 16.002 ms 15.546 ms 15.878 ms
8 196.31.63.202 (196.31.63.202) 15.855 ms 15.534 ms 15.714 ms
9 vl9-ah-ha-2.za.mtnbusiness.net (196.30.42.133) 16.017 ms 16.685 ms 15.764 ms
10 cms-gm.afrihost.com (197.242.144.102) 15.572 ms 15.893 ms 15.514 ms

They now look like this, for over a year everything was perfect?

Traceroute output:
1 105-237-134-1-esr-lo.mtnbusiness.co.za (105.237.134.1) 4.977 ms 4.833 ms 4.558 ms
2 41.181.221.213 (41.181.221.213) 62.919 ms 91.475 ms 113.804 ms
3 41.181.221.214 (41.181.221.214) 86.399 ms 43.653 ms 43.348 ms
4 rb-dca-2.za--qux-b.za.mtnbusiness.net (41.181.165.118) 37.348 ms 48.771 ms 40.217 ms
5 rb-cr-1.za--rb-dca-2.mtnns.net (196.44.31.122) 37.944 ms 38.425 ms 37.678 ms
6 196.44.31.97 (196.44.31.97) 43.345 ms 43.061 ms 55.625 ms
7 196.31.220.12 (196.31.220.12) 37.929 ms 61.309 ms 44.202 ms
8 196.31.220.25 (196.31.220.25) 45.622 ms 38.197 ms 37.077 ms
9 196.31.63.206 (196.31.63.206) 49.728 ms 44.480 ms 43.347 ms
10 vl9-ah-ha-2.za.mtnbusiness.net (196.30.42.133) 43.928 ms 42.520 ms 43.652 ms
11 cms-gm.afrihost.com (197.242.144.102) 46.500 ms 44.067 ms 64.012 ms

And before you blame my exchange or line, this is what WebAfrica looks likes

Tracing route to is.co.za [196.26.2.114]
over a maximum of 30 hops:

1 * * * Request timed out.
2 8 ms 8 ms 8 ms 196.38.74.246
3 8 ms 7 ms 7 ms cdsl1-umh-gi0-0-0-3102.ip.isnet.net [196.38.74.245]
4 8 ms 8 ms 8 ms 196.26.210.200
5 8 ms 8 ms 8 ms mi-za-umh-p8-ten-0-0-0-1.ip.isnet.net [168.209.93.240]
6 8 ms 8 ms 8 ms mi-za-umh-p7-ten-0-0-0-1.ip.isnet.net [168.209.93.203]
7 18 ms 19 ms 18 ms mi-za-bry-p7-lo0.ip.isnet.net [168.209.255.208]
8 18 ms ms 18 ms core2a-bry-gi1-0-0.ip.isnet.net [168.209.218.2]
9 18 ms 19 ms 18 ms bry.it.is.co.za [196.14.64.36]

AfriGenie has given no real support what so ever, just generic answers..... I have PM'ed him a number of trace routes and tested with his account which make NO difference what so ever! I have logged faults with Telkom and Charl (TelkomZA) has tested and confirmed this a problem on the MTN network and the routes need to be fixed from MTN's side as all other ISP's work fine.

You posted on Friday but this is where it stopped?? 6262

Something very weird is going on here, sounds like the BRAS is mis-configured in the area causing funky routing there.
Could you PM me your DSL username and physical address please?
 
I am on a LAN connection, sitting 2m away from my router. Thing is in that table the 9ms was the highest but the average response time was 0 as per the table, so I really don't think the problem is there.





Okay that I can understand, since it occurred on another ISP, but what can be done to resolve this now? You giving me nice explanations but not many solutions...

Our hands are mostly tied in terms of congestion. It's all run on strict DSLAM upgrade plans by Telkom :(
I can have a fault logged for you, but chances are Telkom will close it citing "scheduled upgrades, no ETR" or something along those lines.

What line sync are you on at the moment?
 
10210, TelkomZA or AH live chat.

Thank you. Was hoping there was a way I could do it myself - similar to what was available on cz.
AH has said that hopefully that functionality will be back in the future.
 
Can't stream youtube at 480p while downloading torrent right now. Never had this problem before.

Hell even browsing is slow while downloading a torrent.
 
according to my preferences 100 peers per torrent. Got few running. I've got few running. Not sure if all have 100 peers but I'm assuming that because there's few torrents running simultaneously that might be causing saturation on the line?
 
according to my preferences 100 peers per torrent. Got few running. I've got few running. Not sure if all have 100 peers but I'm assuming that because there's few torrents running simultaneously that might be causing saturation on the line?

Most likely, yes. 100 peers is essentially 100 different connections pulling content down.
A saturated line normally has high latency to the exchange. Similar to when you use Dropbox - my home line shoots up to 1500ms on the 2nd hop when I use Dropbox for uploading a lot of content.
 
Status
Not open for further replies.
Top
Sign up to the MyBroadband newsletter
X