Acceptable Speeds on a FTTH service?

Over-Dose

Well-Known Member
Joined
Mar 15, 2010
Messages
223
Reaction score
2
Location
Cape Town
I'm wondering what the acceptable speed variances are for a fibre line that's operating optimally, and also when/how to know that there IS an issue, purely based on rudimentary speed tests.

I've been on an Octotel/Vox FTTH line for about 2 years now and was "upgraded" (with price increases sadly) from 50/50 to 100/100 and now, with recent Octotel price hikes, am on a 150/150 line.

The 50/50 & 100/100 would have marginal differences when conducting speed tests (via Speedtest.net and Fast.com) which tested (lower) by between 5 - 8Mbps depending on the day, and always tested from a computer connected via ethernet cable to the router.

Now that the 150/150 speed has been availed, speed tests show a greater variance between the stated speeds versus the actual speeds. At a different location a 100/100 line (on Vuma/Vox) tests constantly near 100/100 even when other users are active on the line.

So I'm wondering if what is acceptable, and not worth logging a support ticket over, or whether there are additional ways to determine whether a residential fibre line has issues. I'm aware of other tests like packet loss etc and am not sure how applicable they are.

My current speed test via a 150/150 line, with no other devices active, using the Vox default MirkoTik hAP ac² router:
Vox Speedtest_27.02.2024.png
I am aware of the current Octotel outages, and so far, it appears to not have affected connectivity at all, but perhaps may be the reason for the variance seen via the tests.

Any thoughts would be appreciated.
 
Hi, is this using a network cable to test with?
 
I'm wondering what the acceptable speed variances are for a fibre line that's operating optimally, and also when/how to know that there IS an issue, purely based on rudimentary speed tests.

I've been on an Octotel/Vox FTTH line for about 2 years now and was "upgraded" (with price increases sadly) from 50/50 to 100/100 and now, with recent Octotel price hikes, am on a 150/150 line.

The 50/50 & 100/100 would have marginal differences when conducting speed tests (via Speedtest.net and Fast.com) which tested (lower) by between 5 - 8Mbps depending on the day, and always tested from a computer connected via ethernet cable to the router.

Now that the 150/150 speed has been availed, speed tests show a greater variance between the stated speeds versus the actual speeds. At a different location a 100/100 line (on Vuma/Vox) tests constantly near 100/100 even when other users are active on the line.

So I'm wondering if what is acceptable, and not worth logging a support ticket over, or whether there are additional ways to determine whether a residential fibre line has issues. I'm aware of other tests like packet loss etc and am not sure how applicable they are.

My current speed test via a 150/150 line, with no other devices active, using the Vox default MirkoTik hAP ac² router:
View attachment 1667559
I am aware of the current Octotel outages, and so far, it appears to not have affected connectivity at all, but perhaps may be the reason for the variance seen via the tests.

Any thoughts would be appreciated.
Ya thats a difficult one. I have the same thing going here. I have a 200/200 line with Herotel. When I do a speed test I get 188/187. Always at any time of the day, that's the speed I get. So I phoned Herotel explaining to them what I noticed. And they responded saying. Because speedtest.net is not that accurate. I have to look under Network inside the Task Manager. There you will see the correct speed that you are getting. So I opened my Task Manager, went to Network and did a speedtest. My Network showed me I am getting 198-199.
So I recommend and just check your Network what you are getting. I don't like it like this. I prever getting 200 on my url speed test. But technically what Herotel explained makes sense.
 
Ya thats a difficult one. I have the same thing going here. I have a 200/200 line with Herotel. When I do a speed test I get 188/187. Always at any time of the day, that's the speed I get. So I phoned Herotel explaining to them what I noticed. And they responded saying. Because speedtest.net is not that accurate. I have to look under Network inside the Task Manager. There you will see the correct speed that you are getting. So I opened my Task Manager, went to Network and did a speedtest. My Network showed me I am getting 198-199.
So I recommend and just check your Network what you are getting. I don't like it like this. I prever getting 200 on my url speed test. But technically what Herotel explained makes sense.
Thanks for this tip, it's a very interesting/different way of checking speed tests, since I've always reverted to using Fast.com to verify what my actual streaming speeds would be, and speedtest.net would then be what speeds I can locally.

So I tested via Task Manager it matches what the online speed tests show, so essentially my speed test (as per the screenshot in the original post) is the same as what Task Manager shows.

on Afrihost with a 250Mbps line I consistently get 260Mbps via my WiFi mesh system
This is what I would expect also, since unlike other internet types (e.g. LTE or other Wireless services), fibre has always been promoted as being more stable than say LTE which can have a huge variance in what the consumer experiences.
 
My Dad is on RSAweb paying for 200Mbps. Every day they downgrade him to 100Mbps. I have to phone up pretend i'm plugged in with a cat 6 cable, standing on one leg etc etc then eventually they reset the line and it goes back to 200mbps for the rest of that day.
Seemingly the "...standing on one leg etc etc..." approach from tech support consultants is a default when they're "troubleshooting". Most non-tech-savvy customers are in serious trouble if this is the standard.
 
I'm wondering what the acceptable speed variances are for a fibre line that's operating optimally, and also when/how to know that there IS an issue, purely based on rudimentary speed tests.

I've been on an Octotel/Vox FTTH line for about 2 years now and was "upgraded" (with price increases sadly) from 50/50 to 100/100 and now, with recent Octotel price hikes, am on a 150/150 line.

The 50/50 & 100/100 would have marginal differences when conducting speed tests (via Speedtest.net and Fast.com) which tested (lower) by between 5 - 8Mbps depending on the day, and always tested from a computer connected via ethernet cable to the router.

Now that the 150/150 speed has been availed, speed tests show a greater variance between the stated speeds versus the actual speeds. At a different location a 100/100 line (on Vuma/Vox) tests constantly near 100/100 even when other users are active on the line.

So I'm wondering if what is acceptable, and not worth logging a support ticket over, or whether there are additional ways to determine whether a residential fibre line has issues. I'm aware of other tests like packet loss etc and am not sure how applicable they are.

My current speed test via a 150/150 line, with no other devices active, using the Vox default MirkoTik hAP ac² router:
View attachment 1667559
I am aware of the current Octotel outages, and so far, it appears to not have affected connectivity at all, but perhaps may be the reason for the variance seen via the tests.

Any thoughts would be appreciated.
With Vox you wasting your time they going to just hit you with the service is UP TO 150.

Octotel network sadly is not the most robust, we have one at work and it performs up and now and is never consistent. Some days through IPERF tests i can see packet loss other days its fine.

My home line via Vumatel and Cool Ideas always runs at 105/105 roughly im always just slightly over the 100 mark
 
With Vox you wasting your time they going to just hit you with the service is UP TO 150.

Octotel network sadly is not the most robust, we have one at work and it performs up and now and is never consistent. Some days through IPERF tests i can see packet loss other days its fine.

My home line via Vumatel and Cool Ideas always runs at 105/105 roughly im always just slightly over the 100 mark
I see. This is the kind of feedback that helps. I will admit I haven't as yet run the Iperf tests to check for packet loss (I didn't know it was a thing in fact), but I may just do a bit of reading and then do tests to post on a thread I saw on here to see how it performs.
 
I see. This is the kind of feedback that helps. I will admit I haven't as yet run the Iperf tests to check for packet loss (I didn't know it was a thing in fact), but I may just do a bit of reading and then do tests to post on a thread I saw on here to see how it performs.
A quick fire test to see for packet loss is doing international speedtests to servers that peer directly in UK for example. The fact you not getting 150/150 local would suggest there is some loss already. As you go further the speed will be reduced more as the loss gets bigger.
 
With Vox you wasting your time they going to just hit you with the service is UP TO 150.
100% this.
Vox loves quoting their T&Cs which states "Sorry sir, we can't guarantee international traffic". The amount of times I heard that line get repeated still enrages me to this day.
 
My Dad is on RSAweb paying for 200Mbps. Every day they downgrade him to 100Mbps. I have to phone up pretend i'm plugged in with a cat 6 cable, standing on one leg etc etc then eventually they reset the line and it goes back to 200mbps for the rest of that day.

This may be the ONT? And a port getting stuck at 100Mbps somewhere.

The ISP doesn't control speeds via the line. The FNO does.
 
I see. This is the kind of feedback that helps. I will admit I haven't as yet run the Iperf tests to check for packet loss (I didn't know it was a thing in fact), but I may just do a bit of reading and then do tests to post on a thread I saw on here to see how it performs.
Pingtest is a simpler program, iPerf is good fun if you want to geek out on packet size and type testing
 
This may be the ONT? And a port getting stuck at 100Mbps somewhere.

The ISP doesn't control speeds via the line. The FNO does.
RSAWEB supply a Mikrotik from what i know and they don't know how to set them up correctly. We had for months issues where our line would negotiate to 25 meg when it should be 50meg business. In the end our IT consultant knew someone at RSAWEB and they left us on uncapped and the reason was they were unsure how to cap the line on 50meg and keeping it at 50. So our local now runs at like 650 and international about 100
 
My Dad is on RSAweb paying for 200Mbps. Every day they downgrade him to 100Mbps. I have to phone up pretend i'm plugged in with a cat 6 cable, standing on one leg etc etc then eventually they reset the line and it goes back to 200mbps for the rest of that day.
That is BS, you should consider changing ISP
 
RSAWEB supply a Mikrotik from what i know and they don't know how to set them up correctly. We had for months issues where our line would negotiate to 25 meg when it should be 50meg business. In the end our IT consultant knew someone at RSAWEB and they left us on uncapped and the reason was they were unsure how to cap the line on 50meg and keeping it at 50. So our local now runs at like 650 and international about 100
Now THIS sounds like a solution a win-win solution I should consider. Ditch Vox and switch to RSAWeb. :cool:
 
Now THIS sounds like a solution a win-win solution I should consider. Ditch Vox and switch to RSAWeb. :cool:
I wouldn't recommend them... this situation went on for like 8 months and they were billing us R5500 per month for a 50/50 line running at 25/25... their support were clueless. When our contract ended they took 6 months to change us to a new price (R2400) so they billed us for those 6 months at R5500 and refused to give us a credit...
 
Top
Sign up to the MyBroadband newsletter
X