It tests the speed to the HUB itself over the VM network to ensure it's not a device related issue.
This does, of course, assume that we blindly trust what the Samknows tests are saying and how reliable they are - now although I have no doubt that the underlying technology and methodology are sound, how well this has been implemented in the firmware on the VM hub is somewhat open to debate no?
Perhaps you can explain how the upload to the device is reporting as being twice the speed to the hub itself, the apparent latency to the hub is some five times more than the latency to the device, and the jitter seems to be some 25 time worse to the hub than to the device!
Difficult to reconcile these, no, while still claiming that the SamKnows test are perfectly valid? And yes I do know that the tests are asynchronous.
All I'm saying is that if the reported figures can be that far out, then surely it is folly to claim that if the SamKnows tests show xyz speed to the router, then all is well and the issue has to be with the customer's own kit.
The speed test isn't completed at the same time, so if the upload speed is inconsistent from one moment to the next that would explain the difference.
Jitter shows the same story... and latency!
To be honest you could say the same thing about any speed test site. Very rarely are they consistent.
Thanks for posting and welcome to the community.
Sorry to hear of the low upload speed. I have ran some checks here and everything is fine. No outages/issues and all your levels are fine. There's no issues with SNR (signal to noise ratio) or congestion. The hub is showing as being online for 6 weeks, could you reboot the hub for me and then do the test again?
Need a helpful hand to show you how to make a payment? Check out our guide - How to pay my Virgin Media bill
As I said Carl, I am perfectly aware that the tests are asynchronous, and your point about any speed test site being inconsistent is perfectly valid.
But the difference is that we could choose any number of different speed test sites and compare them, if four give a broadly similar value but one was way, way different, we could, realistically conclude that something was up with the latter and disregard it, no?
The problem here is that we are required to take it on trust that the SamKnows speed test as implemented by a process running on the hub, is infallible and must be regarded as definitive - there is no way of independently checking what it claims with alternatives.
Now this would be OK if it was just to be regarded as indicative, but my point was, that based on the above examples, the differences between the two values is so wildly varied as to call into question the entire validity of the tests themselves. Again this wouldn't necessarily be a problem except there have been many posts on here from customers with similar issues, being asked to run a SamKnows test from the hub, even being told to put the hub back into router mode to do this, getting a result which 'claims' 1140 Mb/s to the hub and 400 Mb/s to the device, and being told in no uncertain terms that this proves that the hub is getting the full speed and the issue is with your equipment; from VM's perspective, not our problem guv and here's the proof!
My contention is that we simply can't assume that.