Engineer came today and discovered a forward path attenuator in the box outside. Removed that then of course the levels were two high anyway I've got one connected directly to the hub.
I'm surprised that it was there too be honest and I suspect it may have been put on when the engineer came to fix next doors phone and temporarily patched their none working phone into my box as temp fix. Needless to say that temp fix is still there and a rather bodged affair as well. The cables joints were outside the box and the cable too tight to put it inside. It could that the bush was in the way at the time and they were coming back later to do a proper fix and never came back. The cover of the box is now held on with three cable ties. Today's engineer used one but the cover was hanging open so I added to more two to make the fit better.
DSl and Thinkbroad band speedtests on the single thread download are all over the place. Speeds anything from 350 to 80. Speedtest single thread reports okay which makes me think the single thread test doesn't do anything.
Downloading ubuntu iso image 1st time it was downloading at 45mb/sec 2nd time 4.7mb/sec 3rd time 25mb/sec which sort of backs up dsl and thinkbroad erratic speed results.
I wonder if this is due to the bodged cabling on the phone which is tapped into my connection.
With the greatest of respect, why not just use a multithreaded protocol?
Like you say, most people dont notice so its not an issue as far as VM are concerned. Depending on your CMTS and the QOS scheduling/how busy it is it will affect you more/less. Then you need to take into account what frequencies EVERY hub on the segment are using to load balance effectively, my understanding is its a processor cycle VS "obvious detriment to the end user" issue. But sometimes I make stuff up in my head.
I mean im pernickety, but consign this to the "yeah it does that" bin along with several other issues....
The thing that raises flags for me is you should NOT be able to "see" the difference, i.e when browsing. Once the area fault is sorted, revisit and if browsing is still markedly different, likely a circuit impairment....
The other thing i would check is LAN throughput when your browser is baulking out.