Afrihost MTN Fixed LTE-A Thread

[BCR]

I have a MTN Fixed LTE sim also so I do keep an eye out if something goes wrong. There isn't a known issue with data usage so if you do have an issue I would recommend getting an app to monitor and then if you notice any discrepancies let me know so we can investigate.

@AfriNatic

Think I've worked out what it is about downloading on your network.

I notice at the top of my phone (where it shows connection /download speeds), after the download has peaked & finished, it seems to drop from the top speed (for eg. between 45 - 55 Mbps, when working well) down to about 15 Mbps for a few seconds.

If I recall correctly, the others seemed to drop into the hundreds of kbps, so I suppose sitting at around 15 Mbps for a few secs. uses some additional data.

Don't know if that makes sense to you & a worthwhile explanation ?

That said, so far it's more stable (compared to when I got this new SIM, as well as the SIM from 11 months ago) & speeds seem acceptable.

Will need to determine if running through data faster requires upgrading to next package tier, from Jan or Feb 2022.

That would take it to "only" R150 pm less than a 30/3 Frogfoot fibre line, where data usage plays no part (ie. unlimited) ... but Frogfoot in PE doesn't have a good track record, so would have to see about that later on (as it is already available to me).

 

[Pulserider]

@AfriNatic Is there perhaps a way to lock my Huawei B535 router to a Cell or something? What I seen now is that it is jumping between two cell's. One Cell is giving me 50Mbps + and the other cell is giving me less than 20Mbps. For some reason while it is switching between these two cells the connection drops for that time. Is there a way to over come this maybe? Settings etc.?

 

[AfriNatic]

@AfriNatic Is there perhaps a way to lock my Huawei B535 router to a Cell or something? What I seen now is that it is jumping between two cell's. One Cell is giving me 50Mbps + and the other cell is giving me less than 20Mbps. For some reason while it is switching between these two cells the connection drops for that time. Is there a way to over come this maybe? Settings etc.?

I don't believe the router allows for PCI locking. That is what you would require to lock it to a tower.

 

[Pulserider]

At the moment the frequency is auto, will selecting the frequency perhaps help? Like 1800+2100 etc

 

[AfriNatic]

At the moment the frequency is auto, will selecting the frequency perhaps help? Like 1800+2100 etc

You can try that to see if it doesn't change.

 

[Dretastic06]

https://mybroadband.co.za/news/5g/426084-mtn-launches-100mbps-uncapped-5g.html @AfriNatic please do your thing

 

[BCR]

@AfriNatic - some (more) feedback :

Amazed since power restored after the mandatory 2 hours out by municipality, a full 5 bars, immediately after switching it back on & up to now - over 2 hours later.

Cell C & Telkom where I am struggle in this regard, to stabilize.

Download speeds have also been good - into the 70s and 80s on occasion.

But usage figures still a concern ; once my Cell C data's finished (either overnight or night after) I'll be using MTN through yourselves primarily, so can hopefully compare your app's figures against Android app I've downloaded.

My concern isn't this month, as still have the 100 GB bonus data to work with ; it's how fast the normal data will be consumed, from Jan 2022 onwards.

 

[BCR]

" ... only be effective on the 1st of the following month, regardless of when in the month the request is made or whether it is an upgrade or downgrade."

@AfriNatic - to clarify the point above, should I want to upgrade, must it also be done by the 15th of (any given) month?

I think downgrades work like that, so want to be sure / reminded on upgrade due date as well.

Thanks.

 

[AfriNatic]

" ... only be effective on the 1st of the following month, regardless of when in the month the request is made or whether it is an upgrade or downgrade."

@AfriNatic - to clarify the point above, should I want to upgrade, must it also be done by the 15th of (any given) month?

I think downgrades work like that, so want to be sure / reminded on upgrade due date as well.

Thanks.

The 15th is mainly for cancellations but our system is forgiving for clients that do it a few days after that.

Upgrades can be requested anytime before the end of the month and they will be scheduled for the 1st. Remember that roll over data expires when upgrading or downgrading. If an invoice was already generated that invoice will be credited by the system automatically and a new invoice generated for the correct package.

 

[BCR]

"But usage figures still a concern ; once my Cell C data's finished (either overnight or night after) I'll be using MTN through yourselves primarily, so can hopefully compare your app's figures against Android app I've downloaded."

Quote taken from my one previous post.

@AfriNatic :

What I realised is using an Android app on my phone to compare your Afrihost app figures against, would isolate usage figures to my phone only, and when I get to put a 2nd SIM from Telkom in the phone would further mess up the comparison figs.

Therefore, just after midnight on the 7th Dec, I logged into the B525 and reset data usage to "0", took a screenshot of that page (with date & time on laptop showing) + took a screenshot of data set at "0" in the Afri. app as well, at same time.

Yesterday afternoon, did a visual comparison of the two (no screenshot evidence), because the Afri app only showed a consumption of 35 MB more than the Huawei router.

Will leave a few more days & come back with more feedback.

I'll be more than satisfied if the difference is around the 100 MB mark - even within 200 or 300 hundred megabytes will be fine.

 

[AfriNatic]

"But usage figures still a concern ; once my Cell C data's finished (either overnight or night after) I'll be using MTN through yourselves primarily, so can hopefully compare your app's figures against Android app I've downloaded."

Quote taken from my one previous post.

@AfriNatic :

What I realised is using an Android app on my phone to compare your Afrihost app figures against, would isolate usage figures to my phone only, and when I get to put a 2nd SIM from Telkom in the phone would further mess up the comparison figs.

Therefore, just after midnight on the 7th Dec, I logged into the B525 and reset data usage to "0", took a screenshot of that page (with date & time on laptop showing) + took a screenshot of data set at "0" in the Afri. app as well, at same time.

Yesterday afternoon, did a visual comparison of the two (no screenshot evidence), because the Afri app only showed a consumption of 35 MB more than the Huawei router.

Will leave a few more days & come back with more feedback.

I'll be more than satisfied if the difference is around the 100 MB mark - even within 200 or 300 hundred megabytes will be fine.

If you can we can track this by checking data usage and then have you reset the usage tracker on the router let it run a few days and check if the usage we record match the router.

 

[BCR]

If you can we can track this by checking data usage and then have you reset the usage tracker on the router let it run a few days and check if the usage we record match the router.

I see the figures in the app have - during the course of today - been changed from being measured in MB (with 3 numbers after the decimal point) to GB (now only 1 number after the decimal point).

The result is a less accurate way of comparing from the app's figures to the router (which incidentally are 2 decimals).

As things stood at 18h15, the router showed 4.74 GB used since just after midnight on the 7th Dec.

Your app now only shows :

7th Dec - 1.7 GB used
(I've a screenshot that previously indicated 1700.10 MB) ;

8th Dec (so far) - 3.2 GB

Therefore 4.9 GB used per Afrihost app, so, so far, a 160 MB difference after less than 2 full days best I can tell, because of the use only 1 decimal in your app vs a more standard 2 decimals in the router.

I've posted this earlier than previously mentioned, to see if you agree this presents a problem in comparing the figures ?

Why did the Afri app suddenly change from a MB format to GB, on such a non descript day as the 8th of the month ?

Seems so coincidental that we were messaging about data usage on the same day the change took place, almost as if it SHOULD be made harder (for me) to make comparisons & draw conclusions from them.

 

[AfriNatic]

I see the figures in the app have - during the course of today - been changed from being measured in MB (with 3 numbers after the decimal point) to GB (now only 1 number after the decimal point).

The result is a less accurate way of comparing from the app's figures to the router (which incidentally are 2 decimals).

As things stood at 18h15, the router showed 4.74 GB used since just after midnight on the 7th Dec.

Your app now only shows :

7th Dec - 1.7 GB used
(I've a screenshot that previously indicated 1700.10 MB) ;

8th Dec (so far) - 3.2 GB

Therefore 4.9 GB used per Afrihost app, so, so far, a 160 MB difference after less than 2 full days best I can tell, because of the use only 1 decimal in your app vs a more standard 2 decimals in the router.

I've posted this earlier than previously mentioned, to see if you agree this presents a problem in comparing the figures ?

Why did the Afri app suddenly change from a MB format to GB, on such a non descript day as the 8th of the month ?

Seems so coincidental that we were messaging about data usage on the same day the change took place, almost as if it SHOULD be made harder (for me) to make comparisons & draw conclusions from them.

Morning,

Please screengrab the differences and drop me a pm so I can have our Dev team have a look at the issue.

 

[BCR]

Morning,

Please screengrab the differences and drop me a pm so I can have our Dev team have a look at the issue.

Having checked figures approx 2 days ago, and then again a short while ago, they seem to cancel themselves out (between the router & the Afrihost app).

Therefore, (thankfully) as things stand at the moment I don't feel it's necessary to submit any figures.

If things change on this, will post again.

 

[BCR]

Signal strength : still variable, now generally between 4 and 5 out of 5 bars, but was stuck on 4 for most of the last 2 or 3 days.

Doesn't seem to affect download speeds much, and criticality, internet connection not cutting out (like it was in Jan 2021 on this same package), so for now should all be good.

 

[BCR]

Ridiculous ping a short while ago - very early a.m. so quite strange as network shouldn't be busy ; was 25ms the other day when last did a speed-test, so hope it resolves itself soon.

Speed Test – Fast, reliable, and ad-free internet speed test

Speed test result: Latency: 1038ms | Download: 0.01Mbps | Upload: 1.18Mbps | Provider: Neotel Pty Ltd | Test your Internet connection speed with our fast, reliable, and ad-free tool. Measure your Internet speed now.

speedtest.co.za

Ping: 102ms
Download: 37.97Mbps
Upload: 45.17Mbps

Air Mobile – 2021-12-28 06:29:25
Server: ZA - Cape Town

EDIT :
Identical ping result after router reboot - https://speedtest.co.za/result/5733445

Ping 102ms
Download 31.52Mbps
Upload 44.81Mbps
Air Mobile – 2021-12-28 06:45:13
Server: ZA - Cape Town


2nd EDIT : test using MyBB speed-test ; it immediately started out at 99ms then dropped to 63ms so I don't know if it can be put down to improvement / routing / to another CPT server that makes a difference ? :

MyBroadband Speed Test

Speed test result: Latency: 63ms | Download: 42.86Mbps | Upload: 44.56Mbps | Provider: Air Mobile

speedtest.mybroadband.co.za

Ping: 63ms
Download: 42.86Mbps
Upload: 44.56Mbps

2021-12-28 06:49:00
Server: Cape Town

@AfriNatic - network status in the app. currently for "Fixed Wireless" states "Everything is looking good" so can you determine (later on) if this is / has been likely a localized thing or wider-spread on MTN, perhaps ?

Thanks.

 

[BCR]

Ridiculous ping a short while ago - very early a.m. so quite strange as network shouldn't be busy ; was 25ms the other day when last did a speed-test, so hope it resolves itself soon.

Speed Test – Fast, reliable, and ad-free internet speed test

Speed test result: Latency: 1038ms | Download: 0.01Mbps | Upload: 1.18Mbps | Provider: Neotel Pty Ltd | Test your Internet connection speed with our fast, reliable, and ad-free tool. Measure your Internet speed now.

speedtest.co.za

Ping: 102ms
Download: 37.97Mbps
Upload: 45.17Mbps

Air Mobile – 2021-12-28 06:29:25
Server: ZA - Cape Town

EDIT :
Identical ping result after router reboot - https://speedtest.co.za/result/5733445

Ping 102ms
Download 31.52Mbps
Upload 44.81Mbps
Air Mobile – 2021-12-28 06:45:13
Server: ZA - Cape Town


2nd EDIT : test using MyBB speed-test ; it immediately started out at 99ms then dropped to 63ms so I don't know if it can be put down to improvement / routing / to another CPT server that makes a difference ? :

MyBroadband Speed Test

Speed test result: Latency: 63ms | Download: 42.86Mbps | Upload: 44.56Mbps | Provider: Air Mobile

speedtest.mybroadband.co.za

Ping: 63ms
Download: 42.86Mbps
Upload: 44.56Mbps

2021-12-28 06:49:00
Server: Cape Town

@AfriNatic - network status in the app. currently for "Fixed Wireless" states "Everything is looking good" so can you determine (later on) if this is / has been likely a localized thing or wider-spread on MTN, perhaps ?

Thanks.

It appears that Afrihost support on this forum is selective (as even with mentioning the rep. by name yesterday with some questions involved, my post was ignored).

Presumably all he has time for is running after fibre speed-tests & analysis thereof etc.

Not to worry yourself if you bother reading this ; when the time comes for me to move on, which may be very soon, the so-called "excellent support" others get here won't be at the top of my list, nor on it for that matter, re consideration for sticking around.

Not that you could care, it seems.

 

[AfriNatic]

It appears that Afrihost support on this forum is selective (as even with mentioning the rep. by name yesterday with some questions involved, my post was ignored).

Presumably all he has time for is running after fibre speed-tests & analysis thereof etc.

Not to worry yourself if you bother reading this ; when the time comes for me to move on, which may be very soon, the so-called "excellent support" others get here won't be at the top of my list, nor on it for that matter, re consideration for sticking around.

Not that you could care, it seems.

Morning,

Sorry I missed your post. I get a lot of tags and messages everyday and if I'm off on weekends there is a quite a backlog for me to catch up with and I miss tags sometimes. I try to search for all the Afrihost posts but sometimes I miss them. No ill intent at all.

Have you tried running a continuous MTR with ping plotter to see where the latency is happening?

There hasn't been any general fixed LTE issues and that is why the network status is all clear as we are unable to keep it up to date with specific area outages as MTN doesn't provide us with these reports.

 

[BCR]

Morning,

Sorry I missed your post. I get a lot of tags and messages everyday and if I'm off on weekends there is a quite a backlog for me to catch up with and I miss tags sometimes. I try to search for all the Afrihost posts but sometimes I miss them. No ill intent at all.

Have you tried running a continuous MTR with ping plotter to see where the latency is happening?

There hasn't been any general fixed LTE issues and that is why the network status is all clear as we are unable to keep it up to date with specific area outages as MTN doesn't provide us with these reports.

Okay, thanks @AfriNatic - I appreciate the reply you gave.

I've managed to produce the following MTR reports (clueless as to what they represent), but had thought previously that these were pretty-much meaningless on LTE due to the nature of it being wireless, compared to fibre whereby the PC is connected to the ONT so much more reliable a test ?

The result - after putting the following into the CLI -
mtr -o "LSDR NBAW JMXI" -rw -c 50 google.com >mtr-report-google

Start: 2021-12-29T10:52:56+0200
HOST: XXXXXX Loss% Snt Drop Rcv Last Best Avg Wrst Jttr Javg Jmax Jint
1.|-- homerouter.cpe 0.0% 50 0 50 1.0 0.9 1.0 2.1 0.1 0.1 1.1 2.1
2.|-- ??? 100.0 50 50 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
3.|-- ??? 100.0 50 50 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
4.|-- ??? 100.0 50 50 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
5.|-- 100.127.3.16 0.0% 50 0 50 25.1 21.3 27.4 134.1 2.7 7.9 111. 119.7
6.|-- 100.127.3.17 0.0% 50 0 50 22.9 21.4 24.1 43.4 0.7 2.9 20.3 48.6
7.|-- 100.127.2.182 0.0% 50 0 50 41.6 39.8 41.9 48.7 0.4 1.5 8.3 20.1
8.|-- 72.14.216.130 0.0% 50 0 50 40.7 40.2 42.4 51.8 0.2 1.9 9.1 24.0
9.|-- 172.253.65.161 0.0% 50 0 50 41.5 39.3 43.2 64.5 0.2 2.8 23.7 50.7
10.|-- 142.251.51.219 0.0% 50 0 50 44.8 40.3 44.5 89.8 3.0 4.5 48.0 102.3
11.|-- jnb02s11-in-f14.1e100.net 0.0% 50 0 50 40.6 39.4 41.3 53.6 0.0 1.8 14.0 28.2

Ping has improved gradually since I posted yest. to where it was before ; here's the latest speed-test :

Speed Test – Fast, reliable, and ad-free internet speed test

Speed test result: Latency: 30ms | Download: 7.15Mbps | Upload: 2.03Mbps | Provider: Wireless Professionele Netwerke | Test your Internet connection speed with our fast, reliable, and ad-free tool. Measure your Internet speed now.

speedtest.co.za

ISP: Air Mobile
Ping: 24ms
Download: 28.6Mbps
Upload: 43.74Mbps
Air Mobile – 2021-12-29
10:55:55
Server: ZA - Cape Town

Lower d/load speed indicates congestion at the tower, I presume ?

I subsequently did an MTR test with a LAN (?) cable plugged in from router to the laptop, but the internet is never typically used like this for our LTE, so it's more academic to compare the two results :

Wired connection :

Start: 2021-12-29T11:16:26+0200
HOST: Loss% Snt Drop Rcv Last Best Avg Wrst Jttr Javg Jmax Jint
1.|-- homerouter.cpe 0.0% 50 0 50 0.6 0.5 0.6 0.7 0.0 0.0 0.1 0.5
2.|-- ??? 100.0 50 50 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
3.|-- ??? 100.0 50 50 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
4.|-- ??? 100.0 50 50 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
5.|-- 100.127.3.16 0.0% 50 0 50 20.0 19.8 22.4 47.0 2.8 2.0 25.3 18.4
6.|-- 100.127.3.17 0.0% 50 0 50 22.9 20.1 21.7 29.4 1.2 1.1 8.4 18.7
7.|-- 100.127.2.182 0.0% 50 0 50 40.6 39.1 42.6 99.1 0.8 4.6 58.2 48.7
8.|-- 72.14.216.130 0.0% 50 0 50 39.8 39.1 42.5 57.1 0.3 3.7 16.7 47.2
9.|-- 172.253.65.163 0.0% 50 0 50 42.8 40.0 42.4 56.1 1.4 3.0 15.7 38.5
10.|-- 172.253.65.251 0.0% 50 0 50 40.6 39.9 42.6 56.6 2.8 2.6 14.1 34.5
11.|-- jnb02s03-in-f14.1e100.net 0.0% 50 0 50 42.6 39.1 40.7 43.2 1.2 0.9 3.1 15.5

Don't know if this makes much difference, as unsure how to analyse the figures.
(sorry rows are out of alignment - formatting suggestion/s for future ?).

Can you pls. explain if these MTR tests are worthwhile generally / what they signify, along with whether the flags are appropriate ?


EDIT : I now see that "Last Best Avg Wrst" must relate to ping (latency) values, and don't look too good, especially compared to the result/s speed-tests usually give.
Then Jitter for "Jttr Javg Jmax Jint" - some decent low values, mixed with some terribly high ones.
Can correct me if I'm wrong on those assumptions.

Still don't know how on LTE testing like this would be appropriate vs. fibre or even ADSL / VDSL.

 

[AfriNatic]

Okay, thanks @AfriNatic - I appreciate the reply you gave.

I've managed to produce the following MTR reports (clueless as to what they represent), but had thought previously that these were pretty-much meaningless on LTE due to the nature of it being wireless, compared to fibre whereby the PC is connected to the ONT so much more reliable a test ?

The result - after putting the following into the CLI -
mtr -o "LSDR NBAW JMXI" -rw -c 50 google.com >mtr-report-google



Ping has improved gradually since I posted yest. to where it was before ; here's the latest speed-test :

Speed Test – Fast, reliable, and ad-free internet speed test

Speed test result: Latency: 30ms | Download: 7.15Mbps | Upload: 2.03Mbps | Provider: Wireless Professionele Netwerke | Test your Internet connection speed with our fast, reliable, and ad-free tool. Measure your Internet speed now.

speedtest.co.za

ISP: Air Mobile
Ping: 24ms
Download: 28.6Mbps
Upload: 43.74Mbps
Air Mobile – 2021-12-29
10:55:55
Server: ZA - Cape Town

Lower d/load speed indicates congestion at the tower, I presume ?

I subsequently did an MTR test with a LAN (?) cable plugged in from router to the laptop, but the internet is never typically used like this for our LTE, so it's more academic to compare the two results :

Wired connection :



Don't know if this makes much difference, as unsure how to analyse the figures.
(sorry rows are out of alignment - formatting suggestion/s for future ?).

Can you pls. explain if these MTR tests are worthwhile generally / what they signify, along with whether the flags are appropriate ?


EDIT : I now see that "Last Best Avg Wrst" must relate to ping (latency) values, and don't look too good, especially compared to the result/s speed-tests usually give.
Then Jitter for "Jttr Javg Jmax Jint" - some decent low values, mixed with some terribly high ones.
Can correct me if I'm wrong on those assumptions.

Still don't know how on LTE testing like this would be appropriate vs. fibre or even ADSL / VDSL.

Might be better to just take a screenshot of the results rather than copy and paste to keep the format.

There will always be jitter on wireless as the CPE compensate for signal noise which is always changing.

When it comes to towers and their bandwidth it's a difficult thing to predict and to say what speeds are acceptable and which are not. As you know Spectrum is expensive and networks just don't have enough vs demand. Tower bandwidth for this reason is a finite resource and it's shared between users of that tower. You might find that 10 x people are all downloading large files as the same time and that will drag the average speed of each user down until the heavy users are down with their file transfers.

That's just how bandwidth is shared. You won't get a situation where 1 user will be able to all the bandwidth but high data users drags the average speeds shared between users down so speed can vary significantly.

Towers also face issues with power and backhaul. MTN has an extensive fibre backhaul for their towers but if that goes down the tower might fall back to microwave links which aren't as fast which will reduce the available bandwidth even more.

From the trace routes provided it's difficult to determine if a tower is congested as MTn filters those IP's out not to respond and the first IP that responds is a switch on our network.