Bad local ping afrihost isp/metrofibre line

odanaids

Member
Joined
Oct 28, 2014
Messages
13
Reaction score
3
I'm currently living in Pietermaritzburg and unfortunately, the only fibre lines are with Metro.

I have 1 friend who is 60km away in Westville, and 1 in Joburg who showed me there lines

Westville ping using VOX / ookla 1ms
JHB ping using cool ideas ? ookla 2ms
My ping using Afrihhost/ 10ms

I'm just asking to find out who is at fault here Metro will blame Afrihost and vice versa.

I just want answers

when playing on international UK/Germany its always over 200ms
i have to use mudfish just to get it back down to 180


any help much appreciated
 

Attachments

  • afrihost.png
    afrihost.png
    50.2 KB · Views: 34
  • Cool ideas.png
    Cool ideas.png
    44 KB · Views: 27
  • vox.png
    vox.png
    78.5 KB · Views: 30
  • wintmr.png
    wintmr.png
    21.3 KB · Views: 33
++ @Afrigirl

@odanaids This could likely be a MFN backhaul issue. Reminds me of the time one of our clients had issue with latency (In March) from Umbilo where MFN resolved the fault. Issue ended up being one of the redundant backhaul links being higher latency(due to maintenance), causing the ping fluctuations from 2ms to 10+ms, which I do see in your MTR test there as well
1755603177053.png


So my theory = MFN backhaul from PMB is at fault

it did take us 3 months to resolve the problem with MFN, so I hope your issue doesn't take as long
 
I'm currently living in Pietermaritzburg and unfortunately, the only fibre lines are with Metro.

I have 1 friend who is 60km away in Westville, and 1 in Joburg who showed me there lines

Westville ping using VOX / ookla 1ms
JHB ping using cool ideas ? ookla 2ms
My ping using Afrihhost/ 10ms

I'm just asking to find out who is at fault here Metro will blame Afrihost and vice versa.

I just want answers

when playing on international UK/Germany its always over 200ms
i have to use mudfish just to get it back down to 180


any help much appreciated
Hi.

Has a fault been logged?
You may share a ticket reference please.
 
Hi @Afrigirl not logged this issue as yet this has been present for over a year,
I now just want this to be fixed for good,

I will open a ticket now and share the ref
 
Hi @Afrigirl not logged this issue as yet this has been present for over a year,
I now just want this to be fixed for good,

I will open a ticket now and share the ref
You do not have to log it, you can PM me your email address.
 
you can see the international ping is high, the local ping of 8ms might not be an issue according to you, but this is the main issue for me
 

Attachments

  • image.jpg
    image.jpg
    186.7 KB · Views: 33
you can see the international ping is high, the local ping of 8ms might not be an issue according to you, but this is the main issue for me
Which tool are you using for these tests? If its an AWS service then its out of your ISPs control. What does latency to queen.cisp.co.za look like for instance?
 
Which tool are you using for these tests? If its an AWS service then its out of your ISPs control. What does latency to queen.cisp.co.za look like for instance?
WinMTR

I don't know much about this stuff, I'm not an expert, but the second hop looks like the problem at least to me
coolescalator 8 posts above, pointed out what I was seeing Aswell.

apparently, I'm having dying gasp alarms, I have no idea what that means I used a multimeter on the plug and it's a consistent 11.93v 0mv dips, I mean shouldn't a tech come over and do tests if there are these confirmed gasp alarms?

1757012813672.png

below was directly from ONT

ont.png
 
WinMTR

I don't know much about this stuff, I'm not an expert, but the second hop looks like the problem at least to me
coolescalator 8 posts above, pointed out what I was seeing Aswell.

apparently, I'm having dying gasp alarms, I have no idea what that means I used a multimeter on the plug and it's a consistent 11.93v 0mv dips, I mean shouldn't a tech come over and do tests if there are these confirmed gasp alarms?

View attachment 1845676

below was directly from ONT

View attachment 1845689
170ms to London from KZN is normal, KZN to JHB ~10ms and then 160ms to London.

Those tests you did with the graphic and the cities isnt Winmtr?
 
WinMTR

I don't know much about this stuff, I'm not an expert, but the second hop looks like the problem at least to me
coolescalator 8 posts above, pointed out what I was seeing Aswell.

apparently, I'm having dying gasp alarms, I have no idea what that means I used a multimeter on the plug and it's a consistent 11.93v 0mv dips, I mean shouldn't a tech come over and do tests if there are these confirmed gasp alarms?

View attachment 1845676

below was directly from ONT

View attachment 1845689
If your speeds are fine and you have no loss they not realistically going to do much and just say it is a higher overhead. I had a friend in Centurion who was with Metro on a backhaul with high latency and they straight up said your line is running fine, speeds are fine and there is no loss
 
If your speeds are fine and you have no loss they not realistically going to do much and just say it is a higher overhead. I had a friend in Centurion who was with Metro on a backhaul with high latency and they straight up said your line is running fine, speeds are fine and there is no loss
I would like to know if switching to a different isp would fix this issue, or is it the metrofibre line itself, and they all just share the same paths.

Im pretty sure if the ONT was on fire and I told them, they would say plug it off for 5 min and reconnect it.
170ms to London from KZN is normal, KZN to JHB ~10ms and then 160ms to London.

Those tests you did with the graphic and the cities isnt Winmtr?

Your looking at the best ping, it's the same way my best local ping is sometimes 2ms but 90% of the time it's actually 8ms, imagine your playing a game on Germany severs which majority are hosted, everyone of your friends are getting 178-180 constant smooth, and the whole time your going 190-200 then 210 then back to 190 then 200, oh awesome I got 182 for 3 seconds oh **** back to 210 woah it's now 280 that's the experience.
 
I would like to know if switching to a different isp would fix this issue, or is it the metrofibre line itself, and they all just share the same paths.

Im pretty sure if the ONT was on fire and I told them, they would say plug it off for 5 min and reconnect it.


Your looking at the best ping, it's the same way my best local ping is sometimes 2ms but 90% of the time it's actually 8ms, imagine your playing a game on Germany severs which majority are hosted, everyone of your friends are getting 178-180 constant smooth, and the whole time your going 190-200 then 210 then back to 190 then 200, oh awesome I got 182 for 3 seconds oh **** back to 210 woah it's now 280 that's the experience.
Yeah so the local ping is determined by the FNO (if it is your first hop)

Only thing to sway that is to change FNO, Openserve not an alternative by chance?
 
No unfortunately Pietermaritzburg is all metro basically the entire area, when I first arrived I was given 1 month of 500/500 think it's a afrihost thing not sure, and then it was 5ms local, so I'm not sure if that was a non congested line Or a priority line. As soon as it was reverted to my normal 250/250 line 10ms.
 
No unfortunately Pietermaritzburg is all metro basically the entire area, when I first arrived I was given 1 month of 500/500 think it's a afrihost thing not sure, and then it was 5ms local, so I'm not sure if that was a non congested line Or a priority line. As soon as it was reverted to my normal 250/250 line 10ms.
It may be the way Metrofibre load balances the services, again this isnt something within your ISPs control.

MFN hands over the service to us at layer2, we have no control of that. Based on your trace it looks like 3-7ms average to your gateway?

But it goes all the way up to 30ms, so maybe something they need to investigate.
 
I would like to know if switching to a different isp would fix this issue, or is it the metrofibre line itself, and they all just share the same paths.

Im pretty sure if the ONT was on fire and I told them, they would say plug it off for 5 min and reconnect it.


Your looking at the best ping, it's the same way my best local ping is sometimes 2ms but 90% of the time it's actually 8ms, imagine your playing a game on Germany severs which majority are hosted, everyone of your friends are getting 178-180 constant smooth, and the whole time your going 190-200 then 210 then back to 190 then 200, oh awesome I got 182 for 3 seconds oh **** back to 210 woah it's now 280 that's the experience.
In my friends case she was going through Metro as an ISP and her local ping was terrible, their support was really bad and i suggested she move to CISP at the time and funny enough when she moved the ping went to normal.

It could have been a a timing thing though like as she moved it got fixed
 
In my friends case she was going through Metro as an ISP and her local ping was terrible, their support was really bad and i suggested she move to CISP at the time and funny enough when she moved the ping went to normal.

It could have been a a timing thing though like as she moved it got fixed
Well with MFN you have to pay for different NNI capacities, potentially if an ISP doesnt upgrade their NNI it can happen I suppose.

But then it would be very much time based.
 
Well with MFN you have to pay for different NNI capacities, potentially if an ISP doesnt upgrade their NNI it can happen I suppose.

But then it would be very much time based.
Ah yes this jogged my memory her ping jumped during peak times and they always did tests during the day and said it was fine
 
Top
Sign up to the MyBroadband newsletter
X