Cool Ideas Fibre ISP - Feedback

[Orihalcon]

Still offline...

Any ETA untill this is fully resolved?

 

[Tinuva]

How is that possible? If the vumatel distribution unit you connect to have power failure then no amount of inverters at home will keep the network up.

This is exactly why home off grid power is pointless if the main issue is having internet access since internet access is dead external from your power source.

 

[Johnatan56]

2 hours seems a bit excessive.

 

[John Tempus]

That Vumatel response is utter bullshit. They do not have backup power at all "substations"

Case in point, I finally got an IP from my region (after 5hrs) and there is no traffic over the vumatel network aside from 1 out of 100 packets getting through at over 5seconds delay.

Take a look at what Vumatel refer to as backup network power, this is all hitting the vumatel network before any CISP interaction later down before someone want to scream CISP is screwed. Look at that stable network 75% packetloss, must be billion people still trying to sync according to their BS excuse. To me this is very straight forward to explain, MASSIVE sections of vumatel backhaul went down due to no backup power at their own stations and these are likely still down leaving traffic rerouted over valid paths that is now massively congested.

| 154.0.1.125 - 73 | 314 | 85 | 790 | 835 | 893 | 867 |
| 154.0.1.13 - 73 | 314 | 85 | 795 | 836 | 884 | 867 |
| 154.0.1.245 - 74 | 313 | 84 | 802 | 850 | 908 | 907 |
| 154.0.5.145 - 77 | 302 | 70 | 800 | 855 | 910 | 866 |
| 154.0.5.133 - 79 | 297 | 64 | 812 | 856 | 927 | 889 |
| 196.223.14.41 - 73 | 315 | 86 | 0 | 861 | 944 | 862 |
| 72.14.237.239 - 72 | 319 | 91 | 0 | 860 | 907 | 860 |
| 216.58.223.14 - 74 | 311 | 81 | 0 | 852 | 908 | 848 |


The Vumatel comment actually makes me kinda angry thinking their users are total retards and to make matters worse I had active power to my CPE when loadshedding occurred in my area and my fiber was DEAD.

Vumatel please do not take the approach of all your clients are idiots and will accept BS explanations such as that quoted above.

Their analogy about microsecond packets for authentication compared to people storming into a stadium is hilarious, wtf are these people smoking. There is ZERO similarities , how the hell can they compare orchestrated data packets with randomized human interaction.

PS. I only now noticed your image containing Vumatel - Stellenberg, that is my region and it sure as hell went down at loadshedding and is still completely unusable after 5hrs.

 

[Bighit]

2 hours seems a bit excessive.

Well, it was off for 3 hours.

 

[Orihalcon]

Still off. 5 hours and counting.

 

[doubletrouble]

Still off here... Going on 6 hours.

 

[TheRoDent]

Still off. 5 hours and counting.

The full monty.

Dear partner

Please note that Eskom has now announced Stage 4 load-shedding. This may impact your customer and may result in DHCP storms on the Vumatel network. We have included a customer-friendly communication below which you may share with your customers in the event that they require feedback on this specific issue.

The impact of load-shedding on the Vumatel Network

As you are aware, Vumatel hosts the DHCP server internally on the Active Ethernet network. With the introduction of load-shedding on a mass scale, we have seen DHCP broadcast storms after the power has been restored. Whilst we have segregated the network into smaller DHCP domains, the volumes of DHCP requests, given the size of the load-shedding areas, are still considerably high.

These broadcast storms have resulted in some ISP customers (end-users) receiving an updated IP address, in some instances, later than usual. In a worst-case scenario, this has been recorded up to 2 hours after the power has been restored. The sudden volume of DHCP requests results in some DHCP requests being dropped by the Vumatel switches, requiring repeated DHCP requests before the ISP modem receives a response and assigned IP address.

Vumatel is aware of the impact of area-wide load shedding and has put a plan into place to mitigate these challenges. Significant improvements have been achieved in smaller network areas, but in larger network areas the impact is still being felt. Vumatel is also investigating software changes for the Raycore CPE devices which will allow for less aggressive DHCP requests and less load on the network switches, which will allow for further control of these broadcast storms.

IMPORTANT: These DHCP storms DO NOT occur with every load-shedding incident (only those affecting many areas at once) and DO NOT affect all customers. Only a portion of customers will see their service restoration delayed. It is also only related to the Vumatel Active Ethernet Network. It DOES NOT affect the GPON networks. Vumatel closely monitors the network during load-shedding to mitigate the impact as best as possible and is continually working on this to ensure that we find a sustainable solution to the problem.

Suggested communication to the end-user:

How load-shedding can affect your fibre connection

Vumatel's network is backed up by generators and UPS hardware so if you have backup power* at home, you will be able to surf the internet during load-shedding.

Load-shedding is currently occurring on a mass scale across many suburbs at once. This means that at any given time, thousands of Vumatel customers (without back up power) are going down due to a loss of power to their CPE (white fibre Raycore device mounted to their wall). When the power is restored, thousands of CPEs try to restore their connection with the network at once. This can cause a delay for some customers in the restoration of services immediately after load-shedding. It is similar to thousands of people suddenly trying to enter a stadium at the same time and only so many people can fit through the gates at once. If you find that your internet service does not come back up straight after load-shedding, please be patient. The CPE is trying to authenticate a connection, along with all the others that have just come online. If after 2 hours, the service has still not been restored, please contact your ISP for assistance.

*Backup Power Requirements for the CPE & Wi-Fi router:

To keep your internet connection up during Stage 2 load-shedding (up to 4 hours), by installing a 1000Va UPS, you will be able to keep the following devices powered up during this period:

1. Vumatel CPE (white Raycore device)
2. ISP Wi-Fi Router
3. One Mobile device based power supply (1 - 5 amp) which will typically charge smartphones and tablets during this period

Please note: this is an indicative calculation and will vary based on the different Wi-Fi routers and power supplies of mobile devices connected to the UPS during this period.

To keep a television powered during this the recommended size of the UPS would be no less than 3000Va.
--
Kind Regards,
The VUMA Team

 

[Orihalcon]

Thank you.

Access is 'restored' but still can't really do anything.

Oh well. Will have to invest in some UPS units.

 

[TheRoDent]

Thank you.

Access is 'restored' but still can't really do anything.

Oh well. Will have to invest in some UPS units.

PM me your IP details and network?

 

[Tinuva]

That Vumatel response is utter bullshit. They do not have backup power at all "substations"

Case in point, I finally got an IP from my region (after 5hrs) and there is no traffic over the vumatel network aside from 1 out of 100 packets getting through at over 5seconds delay.

Take a look at what Vumatel refer to as backup network power, this is all hitting the vumatel network before any CISP interaction later down before someone want to scream CISP is screwed. Look at that stable network 75% packetloss, must be billion people still trying to sync according to their BS excuse. To me this is very straight forward to explain, MASSIVE sections of vumatel backhaul went down due to no backup power at their own stations and these are likely still down leaving traffic rerouted over valid paths that is now massively congested.

| 154.0.1.125 - 73 | 314 | 85 | 790 | 835 | 893 | 867 |
| 154.0.1.13 - 73 | 314 | 85 | 795 | 836 | 884 | 867 |
| 154.0.1.245 - 74 | 313 | 84 | 802 | 850 | 908 | 907 |
| 154.0.5.145 - 77 | 302 | 70 | 800 | 855 | 910 | 866 |
| 154.0.5.133 - 79 | 297 | 64 | 812 | 856 | 927 | 889 |
| 196.223.14.41 - 73 | 315 | 86 | 0 | 861 | 944 | 862 |
| 72.14.237.239 - 72 | 319 | 91 | 0 | 860 | 907 | 860 |
| 216.58.223.14 - 74 | 311 | 81 | 0 | 852 | 908 | 848 |


The Vumatel comment actually makes me kinda angry thinking their users are total retards and to make matters worse I had active power to my CPE when loadshedding occurred in my area and my fiber was DEAD.

Vumatel please do not take the approach of all your clients are idiots and will accept BS explanations such as that quoted above.

Their analogy about microsecond packets for authentication compared to people storming into a stadium is hilarious, wtf are these people smoking. There is ZERO similarities , how the hell can they compare orchestrated data packets with randomized human interaction.

PS. I only now noticed your image containing Vumatel - Stellenberg, that is my region and it sure as hell went down at loadshedding and is still completely unusable after 5hrs.

Then explain how my internet was working 100% during loadshedding and after loadshedding?

Unfortunately the inverter is only for the TV, routers ect. My fileserver running my smokeping goes down during loadshedding, but it shows everything 100% after the power came back on:
https://netmon.heaven.za.net/smokeping/smokeping.cgi?target=CISP

 

[initroot]

Well loadshedding ended in plattekloof and vuma trenched now completely offline.

 

[John Tempus]

Then explain how my internet was working 100% during loadshedding and after loadshedding?

Unfortunately the inverter is only for the TV, routers ect. My fileserver running my smokeping goes down during loadshedding, but it shows everything 100% after the power came back on:
https://netmon.heaven.za.net/smokeping/smokeping.cgi?target=CISP

Because your substation probably did have backup power.

What I am trying to point out is that all vumatel substations do not have backup power, that is outright a lie or they all selectively failed during this load shedding.

 

[SeaSickMama]

2 hours seems a bit excessive.

Just think that is a hyper inflated figure encase they need to restart a sections etc

 

[John Tempus]

Just think that is a hyper inflated figure encase they need to restart a sections etc

Actually under estimation, took 5hrs 30mins for things to get back online here.

 

[DeatheCore]

Well loadshedding ended in plattekloof and vuma trenched now completely offline.

I'm in Panorama and our Vumatel trenched fibre is also completely offline. Blue flashing light on the CPE / no dhcp IP address

Edit: Back online around 10:35 PM

 

[Tinuva]

Because your substation probably did have backup power.

What I am trying to point out is that all vumatel substations do not have backup power, that is outright a lie or they all selectively failed during this load shedding.

Whatever the case... I am going to admit defeat to this argument. 1 I don't know the facts. 2 it's not worth the effort. There surely is an explanation, I just don't know what.

 

[DeatheCore]

Although I'm back up and running here, there's definitely some major transit issues going on.

 

[John Tempus]

Ye the transit seems like an aids victim right now :/

You know it will likely survive but it is such hit and miss that you are not sure.

 

[John Tempus]

Whatever the case... I am going to admit defeat to this argument. 1 I don't know the facts. 2 it's not worth the effort. There surely is an explanation, I just don't know what.

Ye I did not want it to be an argument just reflected on my own personal experience/example. Loadshedding kick in , my fiber to area substation went down/had zero traffic/no sync, nada..