Quote:
Originally Posted by roughbeast
I just don't use thinkbroadband for any kind of testing. Although thinkbroadband is ostensibly geared up for 200Mb testing, forinstance, and their ping test is via the same uncongested network as their speed test, their connection to the VM network is very poor. They say they have complained to VM about this, but VM have done nothing. So, their speed test and ping test results probably aren't even worth us discussing if we are on VM.
This result I get day or night with hardly any variation. Who needs Thinkbroadband right now? |
their pings are fine, its their speedtest. Few points I will make here which were made open after alot of pushing from myself and someone else.
1 - ncuk peer over linx on 2 networks, as well as lonap. ncuk host tbb. VM dont peer over lonap so that leaves linx or external transit, ncuk send traffic to VM over linx. Since tbb intially blamed vM they would appear to be saying VM's linx link is congested however ping graphs and other traffic going over linx doesnt back this claim up.
2 - after the pushing it became apparent the bulk of the speedtest issues were down to one of the tbb speedtest server's, when this was bypassed speed's shot up although some issues remained they were much less severe.
Ultimately using the tbb monitor is just a test between 2 end points, so it may not truly reflect generic performance on the connection, but I would say most of the time it should give a reasonable picture of whats going on.