Status
Not open for further replies.
Hi Paul,

Just trying to download the latest Nvidia 397.93 drivers from uk.download.nvidia.com
I am getting 28KB/s will take 5 - 6 hours to download.

Same issue for us.download.nvidia.com


Traceroute shows it goes to the Edgecast CDN via NAPAfrica right here in Joburg.

Am seeing the same kind of slowing down. Initial burst of download, and then slowdown to 30kbps or so. Definately a CDN implemented throttle, not CISP.
 
Have filtered it out of the edges and pulling it international now, seems better?
 
Have filtered it out of the edges and pulling it international now, seems better?

Yes, 4.6 - 5.0MB/s on a single Thread, getting 21MB/s split over 10 Threads.

|------------------------------------------------------------------------------------------|

| WinMTR statistics |

| Host - % | Sent | Recv | Best | Avrg | Wrst | Last |

|------------------------------------------------|------|------|------|------|------|------|

| router.lan - 0 | 10 | 10 | 0 | 0 | 0 | 0 |

| 155.93.246.1 - 0 | 10 | 10 | 1 | 1 | 1 | 1 |

| 154.0.1.125 - 0 | 10 | 10 | 1 | 1 | 1 | 1 |

| 154.0.1.13 - 0 | 10 | 10 | 1 | 1 | 1 | 1 |

| 154.0.3.106 - 0 | 10 | 10 | 141 | 141 | 142 | 141 |

| v1000.core1.lon3.he.net - 0 | 10 | 10 | 142 | 142 | 143 | 142 |

| 100ge14-1.core1.lon2.he.net - 0 | 10 | 10 | 148 | 158 | 166 | 148 |

| 195.66.224.62 - 0 | 10 | 10 | 143 | 143 | 144 | 144 |

| 152.195.97.129 - 0 | 10 | 10 | 142 | 147 | 166 | 144 |

| 192.229.221.58 - 0 | 10 | 10 | 141 | 142 | 143 | 141 |

|________________________________________________|______|______|______|______|______|______|

WinMTR v0.92 GPL V2 by Appnor MSP - Fully Managed Hosting & Cloud Provider
 
Last edited:
Sorry if it has been asked before, searched but with no luck.

Is there a way to get the router from CI before Vuma comes out on the day to install ? I am helping a very difficult guy to get is vuma installed and onto CI, to tell him on Tuesday that he needs to get CI only after the link is active on the Vuma portal and then wait for delivery of the router will cause a nuclear reaction from him...
 
Sorry if it has been asked before, searched but with no luck.

Is there a way to get the router from CI before Vuma comes out on the day to install ? I am helping a very difficult guy to get is vuma installed and onto CI, to tell him on Tuesday that he needs to get CI only after the link is active on the Vuma portal and then wait for delivery of the router will cause a nuclear reaction from him...
Drop me a PM with the details and I will see what I can do.
 
Some strange behavior this morning.

Code:
Tracing route to 8.8.8.8 over a maximum of 30 hops

  1    <1 ms    <1 ms    <1 ms  192.168.0.1
  2    <1 ms    <1 ms    <1 ms  155.93.196.1
  3    <1 ms    <1 ms    <1 ms  154.0.1.121
  4    <1 ms    <1 ms    <1 ms  154.0.4.62
  5    <1 ms    <1 ms    <1 ms  154.0.1.121
  6     *        *        *     Request timed out.

Code:
Tracing route to cisp.co.za [154.0.13.95]
over a maximum of 30 hops:

  1    <1 ms    <1 ms    <1 ms  192.168.0.1
  2     1 ms    <1 ms    <1 ms  155.93.196.1
  3    <1 ms    <1 ms    <1 ms  c3d-backbone.coolideas.co.za [154.0.1.121]
  4     *        *        *     Request timed out.

Using google dns I can access MyBB but most other international sites fail to load.
Can't test any more as I have to go to the office now. :cry:
 
Last edited:
Hi PBCool

It seems like my connection is also down at home. This is in Durban North area. Anything amiss on your side?

Regards
Hellhound
 
Hi PBCool

It seems like my connection is also down at home. This is in Durban North area. Anything amiss on your side?

Regards
Hellhound

Im back up, for now...
I think it is more to do with Vumatel than Cool Ideas.
 
Im back up, for now...
I think it is more to do with Vumatel than Cool Ideas.

It's a fibre break between datacentres, we moved over to a backup link for now whilst we wait for feedback.
 
Out of interest sake, why is the move over to backup link manual process?

The primary path is supposed to be based on a ring service, and so there should be a very slim chance of ever failing.

The backup circuit is a pay per use service. So we can have it automatically fail-over but you risk it running up a massive bill on a false positive situation.

Basically the primary service shouldn't go down, this is the feedback we are awaiting.
 
The primary path is supposed to be based on a ring service, and so there should be a very slim chance of ever failing.

The backup circuit is a pay per use service. So we can have it automatically fail-over but you risk it running up a massive bill on a false positive situation.

Basically the primary service shouldn't go down, this is the feedback we are awaiting.

This is in part why Cool Ideas is building up such a good reputation, you don't treat your customers as "stupids". Even those of us that are less tech savy but still have a keen interest can read and understand what is happening. Or how and why things work. This openness builds trust versus "OTHERS" who used to try baffle our brains with "smoke and mirrors".

When a customer submits an issue you investigate and take action ..... that makes a world of difference versus "Others".
 
[MENTION=394133]PBCool[/MENTION] Are you going to add NAPA traffic again on the graphs?

Yeah we are looking to dump cacti and use something else, which is why we havent updated it with the edge router change.
 
Little bit off topic, but from some past conversations in this thread about windows network stack that is hampering often upload speeds compared to Linux driven devices, well I just read this in the Windows 10 Insiders build 17677 release notes.

“Windows is transforming the networking stack after 20 years through the NetAdapter framework. This framework introduces a new, more reliable, network driver model that inherits the goodness of the Windows driver framework while bringing an accelerated data path. More details coming soon”
 
Last edited:
The primary path is supposed to be based on a ring service, and so there should be a very slim chance of ever failing.

The backup circuit is a pay per use service. So we can have it automatically fail-over but you risk it running up a massive bill on a false positive situation.

Basically the primary service shouldn't go down, this is the feedback we are awaiting.

Thank you for taking the time to explain.
 
Little bit off topic, but from some past conversations in this thread about windows network static that is hampering often upload speeds compared to Linux driven devices, well I just read this in the Windows 10 Insiders build 17677 release notes.

“Windows is transforming the networking stack after 20 years through the NetAdapter framework. This framework introduces a new, more reliable, network driver model that inherits the goodness of the Windows driver framework while bringing an accelerated data path. More details coming soon”

And just yesterday I was wondering if they are ever going to do it. Then I thought, nah, it's MS, they won't. :p
 
Status
Not open for further replies.
Top
Sign up to the MyBroadband newsletter
X