Cool Ideas Fibre ISP - Feedback

[John Tempus]

I don't see any congestion at all during or after loadshedding. I am on the same pop as you. I am 100% certain you are on a switch that is causing that packet loss, where I am on one where I don't see it at all.

Its not the whole network like you think it is.

I am thinking you are one of the lucky few rather than me being one of the unlucky ones.

I am on group of over 200 users in brackenfell/durbanville/stellenberg/kraaifontein , basically the outskirts around where I am and whenever I have these issues they have these issues. Not a single time were our issues not at the same time since Monday. We are most definitely not on the exact same POP nor same switches remotely close to each other.

The issue with DHCP leasing is related to network congested that is caused by what everyone keep calling DHCP Storm.

I know you believe its a switch but my evidence points to it being a broad network issue independent of any specific switch but possibly isolated within certain section of the network in the area I just outlined.

You should fall in the same area, if you claim to not have any of these post loadshedding woes then you are one lucky SOB because its a reality for the majority and 100% of the people I am in communication with.

It does not get clearer than this.

Below using the static ip for reference, during the hours everyone is trying to get IP via DHCP the static ip results look like this.

|------------------------------------------------------------------------------------------|
| WinMTR statistics |
| Host - % | Sent | Recv | Best | Avrg | Wrst | Last |
|------------------------------------------------|------|------|------|------|------|------|
| 192.168.0.1 - 0 | 20 | 20 | 0 | 0 | 0 | 0 |
| 155.93.248.1 - 0 | 20 | 20 | 2 | 12 | 58 | 5 |
| c3h-backbone.coolideas.co.za - 25 | 8 | 6 | 554 | 579 | 602 | 588 |
| cd-backbone.coolideas.co.za - 19 | 11 | 9 | 534 | 569 | 601 | 534 |
| cloudflare.ixp.capetown - 50 | 6 | 3 | 0 | 613 | 628 | 609 |
| 154.0.5.146 - 50 | 6 | 3 | 0 | 613 | 628 | 609 |
|________________________________________________|______|______|______|______|______|______|


Still on the exact same static ip and after the DHCP assignments hours played out ie. I am now actually able to get IP via DHCP, the following is the result still on the same static ip.

|------------------------------------------------------------------------------------------|
| WinMTR statistics |
| Host - % | Sent | Recv | Best | Avrg | Wrst | Last |
|------------------------------------------------|------|------|------|------|------|------|
| 192.168.0.1 - 0 | 55 | 55 | 0 | 0 | 0 | 0 |
| 155.93.248.1 - 0 | 55 | 55 | 2 | 10 | 42 | 21 |
| 154.0.1.125 - 0 | 55 | 55 | 1 | 1 | 2 | 1 |
| 154.0.1.13 - 0 | 55 | 55 | 1 | 1 | 1 | 1 |
| 196.10.140.198 - 0 | 55 | 55 | 1 | 1 | 1 | 1 |
| 1.1.1.1 - 0 | 55 | 55 | 1 | 1 | 3 | 1 |
|________________________________________________|______|______|______|______|______|______|


So the control here is the same active session, obviously the same switch being used and the only difference is during and after DHCP assignments trashing the network.

I am 99.99% if not 100% certain the network is actually getting destroyed due to DHCP assignment, the evidence is crystal clear to me. The only thing that makes no sense to me would be if you have no issue at all ever to get on after loadshedding, you must have some very special wiring.

I have been able to get the exact results on tuesday night and now on wednesday night. Exactly after loadshedding and during hours on end people trying to get back onto the network.

 

[Tinuva]

Not according to the loadshedding schedules, same areas was down earlier the day and then there were no issues when it came back up

Think bigger.

If you only take CoCT schedules into consideration, then yes its the same areas at both times. BUT if you take the whole country's schedules into consideration, you will realize there is a massive difference. Gauteng for example follow 4.5 hour slots, CoCT 2.5 hour slots. Some other provinces have munies following 3 hour slots.

The loadshedding that ens at 9pm or the end of the day, will always be worse than when loadshedding lots end in the middle of the day.

And the reason I say this, if you look at the ip address of the DHCP server, even for me in CT, the DHCP server seems to sit in Gauteng.

 

[KantSnyer]

Switched to Axxess running fine now

 

[Praeses]

How there's still not a notice up regarding the dead international speeds at CISP since last night is beyond me...

 

[semaphore]

How there's still not a notice up regarding the international speeds at CISP is beyond me...

It's pathetic.

 

[John Tempus]

How there's still not a notice up regarding the dead international speeds at CISP since last night is beyond me...

This is a different issue I am sure.

Aside from Vumatel recent problems that start on Monday, I have had zero local or international speed issue. When I finally get back on atm after loadshedding my local and international is perfectly fine also.

You on Vumatel or a different network ?

 

[Viper87]

Yeh its a trench issue. Arial is not managed by Vumatel so finally Arial guys getting good news. lol

Also battling with international this morning, on and then off. Openserve

This seems to happening on 3 different last mile provider connections so there must be something more going on here than any 1 issue across any single network

 

[Tinuva]

I am thinking you are one of the lucky few rather than me being one of the unlucky ones.

I am on group of over 200 users in brackenfell/durbanville/stellenberg/kraaifontein , basically the outskirts around where I am and whenever I have these issues they have these issues. Not a single time were our issues not at the same time since Monday. We are most definitely not on the exact same POP nor same switches remotely close to each other.

The issue with DHCP leasing is related to network congested that is caused by what everyone keep calling DHCP Storm.

I know you believe its a switch but my evidence points to it being a broad network issue independent of any specific switch but possibly isolated within certain section of the network in the area I just outlined.

You should fall in the same area, if you claim to not have any of these post loadshedding woes then you are one lucky SOB because its a reality for the majority and 100% of the people I am in communication with.

It does not get clearer than this.

Below using the static ip for reference, during the hours everyone is trying to get IP via DHCP the static ip results look like this.

|------------------------------------------------------------------------------------------|
| WinMTR statistics |
| Host - % | Sent | Recv | Best | Avrg | Wrst | Last |
|------------------------------------------------|------|------|------|------|------|------|
| 192.168.0.1 - 0 | 20 | 20 | 0 | 0 | 0 | 0 |
| 155.93.248.1 - 0 | 20 | 20 | 2 | 12 | 58 | 5 |
| c3h-backbone.coolideas.co.za - 25 | 8 | 6 | 554 | 579 | 602 | 588 |
| cd-backbone.coolideas.co.za - 19 | 11 | 9 | 534 | 569 | 601 | 534 |
| cloudflare.ixp.capetown - 50 | 6 | 3 | 0 | 613 | 628 | 609 |
| 154.0.5.146 - 50 | 6 | 3 | 0 | 613 | 628 | 609 |
|________________________________________________|______|______|______|______|______|______|


Still on the exact same static ip and after the DHCP assignments hours played out ie. I am now actually able to get IP via DHCP, the following is the result still on the same static ip.

|------------------------------------------------------------------------------------------|
| WinMTR statistics |
| Host - % | Sent | Recv | Best | Avrg | Wrst | Last |
|------------------------------------------------|------|------|------|------|------|------|
| 192.168.0.1 - 0 | 55 | 55 | 0 | 0 | 0 | 0 |
| 155.93.248.1 - 0 | 55 | 55 | 2 | 10 | 42 | 21 |
| 154.0.1.125 - 0 | 55 | 55 | 1 | 1 | 2 | 1 |
| 154.0.1.13 - 0 | 55 | 55 | 1 | 1 | 1 | 1 |
| 196.10.140.198 - 0 | 55 | 55 | 1 | 1 | 1 | 1 |
| 1.1.1.1 - 0 | 55 | 55 | 1 | 1 | 3 | 1 |
|________________________________________________|______|______|______|______|______|______|


So the control here is the same active session, obviously the same switch being used and the only difference is during and after DHCP assignments trashing the network.

I am 99.99% if not 100% certain the network is actually getting destroyed due to DHCP assignment, the evidence is crystal clear to me. The only thing that makes no sense to me would be if you have no issue at all ever to get on after loadshedding, you must have some very special wiring.

I have been able to get the exact results on tuesday night and now on wednesday night. Exactly after loadshedding and during hours on end people trying to get back onto the network.

I think if I had to reset my CPE+Router after loadshedding, I would be stuck in the same situation as everyone else. Not going to take that risk.

I remember in December 2018, before I had my inverter, that it took me 30minutes to get online after loadshedding ended. I very distinctly remember that, but didn't make anything off it at the time. It looks very similar to what everyone is experiencing now...except not as bad.

 

[semaphore]

This is a different issue I am sure.

Aside from Vumatel recent problems that start on Monday, I have had zero local or international speed issue. When I finally get back on atm after loadshedding my local and international is perfectly fine also.

You on Vumatel or a different network ?

You're lucky, because its mot definitely not fine:



And I am not on OpenServe.

 

[irBosOtter]

Think bigger.

If you only take CoCT schedules into consideration, then yes its the same areas at both times. BUT if you take the whole country's schedules into consideration, you will realize there is a massive difference. Gauteng for example follow 4.5 hour slots, CoCT 2.5 hour slots. Some other provinces have munies following 3 hour slots.

The loadshedding that ens at 9pm or the end of the day, will always be worse than when loadshedding lots end in the middle of the day.

And the reason I say this, if you look at the ip address of the DHCP server, even for me in CT, the DHCP server seems to sit in Gauteng.

No difference between load shedding ending 2pm or 9pm, same amount of areas that comes up, they did not add extra areas for later the day if you look at the load shedding schedules for the country

I have zero issues getting a IP address, I reboot 20 times and get IP immediately, but can't even ping the gateway.

 

[John Tempus]

You're lucky, because its mot definitely not fine:

View attachment 619574

And I am not on OpenServe.

Ye there is the problem, different issue that I am not aware of no experience with Openserve but I guess it might be in a different way related to Octotel-like issues specific to certain areas etc etc.

This is the one thing Vumatel users havn't struggled with until now with the Eskom debacles.

 

[Praeses]

This is a different issue I am sure.

Aside from Vumatel recent problems that start on Monday, I have had zero local or international speed issue. When I finally get back on atm after loadshedding my local and international is perfectly fine also.

You on Vumatel or a different network ?

On Frogfoot. My connection was perfect for the first time ever since I got fibre in December (after frogfoot did maintenance yesterday)...then at around 10pm-ish international was dead.

Now I can't even remote desktop into my PC at home to switch it off before loadshedding

 

[semaphore]

Ye there is the problem, different issue that I am not aware of no experience with Openserve but I guess it might be in a different way related to Octotel-like issues specific to certain areas etc etc.

This is the one thing Vumatel users havn't struggled with until now with the Eskom debacles.

This has nothing to do with certain areas, it seems to be when it his some router in the UK to breaks. People are having this from Durban, I am in JHB, they're on OpenServe, I'm on Edge Telecoms.

 

[Viper87]

That's now a 5 Networks this is happening on, across different parts of the country

 

[John Tempus]

No difference between load shedding ending 2pm or 9pm, same amount of areas that comes up, they did not add extra areas for later the day if you look at the load shedding schedules for the country

I have zero issues getting a IP address, I reboot 20 times and get IP immediately, but can't even ping the gateway.

You are lucky this probably means your POP have backup power and doesnt lose your MAC address on lease assignment and simply spit it right back at you each time.

But you point out a interesting aspect , DEAD network with/without IP. Because the rest of the network that is doing DHCP storm DDOSing is killing the network while your first POP that probably have reliable backup power does not get hit by this traffic.

 

[John Tempus]

This has nothing to do with certain areas, it seems to be when it his some router in the UK to breaks. People are having this from Durban, I am in JHB, they're on OpenServe, I'm on Edge Telecoms.

I have no international issue on vumatel, do you have specific place you need to me test so I can confirm ?

To be blunt I can only really speak on what I have personal experience/access with which is Vumatel. Openserve network is designed entirely different with PON that is way easier to congest locally than Vumatel which is openserve negative vs their positves not being affected like us right now during loadshedding.

Gotta weight the pros and cons I guess.

 

[irBosOtter]

You are lucky this probably means your POP have backup power and doesnt lose your MAC address on lease assignment and simply spit it right back at you each time.

But you point out a interesting aspect , DEAD network with/without IP. Because the rest of the network that is doing DHCP storm DDOSing is killing the network while your first POP that probably have reliable backup power does not get hit by this traffic.

My neighbor is on Vumatel and Afrihost ISP, his internet works fine while I have zero, so is that "DHCP storm" only for certain ISP clients then?

 

[John Tempus]

My neighbor is on Vumatel and Afrihost ISP, his internet works fine while I have zero, so is that "DHCP storm" only for certain ISP clients then?

I cant say for certain why you are not on and your neighbor is since you both should be going to the exact same POP. I was only able to comment on my realtime testing using static ip and DHCP ip. Took me 3hrs longer before I was able to get DHCP ip assigned even though the network was functioning using the static ip.

So there might be a possibility that you are still in hypothetical queue to get assigned an IP.

 

[Praeses]

I've opened a ticket regarding the dead international traffic and informed the facebook rep - which seems to be up and responsive.

Ticket number: COOL-20190214-219024

 

[Luffy101]

My international is still buggered. Openserve in Centurion