That particular BT speedtester happens to be the very one that Vispa Support recommended I ran some two weeks ago, when the latency problem first arose and I reported it to them. I didn't actually run it two weeks ago, as I already knew at what speed my connection runs (my 8800NL router gives me all the stats) and I'd already run several ping and traceroute tests. So, do you think that that's an indication that the 'provider' is definitely BTWholesale and not some other outfit?
Regardless, I went to the tester today and, using my Windows machine rather than the Mac I have, I did what you suggested, and the basic test did in fact run. However, for some reason it was impossible to do a copy-and-paste of the results. Possibly, this was caused by my edition of Flash being out of date. It was really annoying because I could highlight everything apart from the very results! Unfortunately, I have my Mac set up to not allow Flash for every single website and it requires my approval for each site I visit. However, this mechanism doesn't work on the bt speedtester.
Thus, I had to just jot down the results. There were only three and they were:
Download speed (Mbps): 3.77
Upload speed (Mbps): 0.63
Ping latency (ms): 54.63
I actually ran this twice, after an interval of about a minute. The second latency time was 60.63.
The two speeds are pretty much correct, and the latency time is about twice as long as I've been seeing for most
hops these last couple of days.
It accepted my wanting to run Further Tests and it started them, but I then got:
"Test Error. The performance tester is currently unable to run a speed test for your broadband connection. Please try again shortly. However, if this problem persists, raise the issue with your service provider".
So, I had to abort the Further Tests.
Aside from the bt speedtester, I've run some traceroutes this morning and have got these results:
Traceroute has started…
traceroute: Warning: www.bbc.co.uk
has multiple addresses; using 22.214.171.124
traceroute to www.bbc.net.uk
(126.96.36.199), 64 hops max, 72 byte packets
1 192.168.1.254 (192.168.1.254) 0.858 ms 0.769 ms 0.786 ms
2 188.8.131.52 (184.108.40.206) 23.003 ms 22.517 ms 22.889 ms
3 edge.td.vispa.net.uk (220.127.116.11) 23.117 ms 23.089 ms 23.107 ms
4 te-8-3-203.car2.manchester1.level3.net (18.104.22.168) 23.583 ms 23.181 ms 23.185 ms
5 ae-122-3508.edge4.london1.level3.net (22.214.171.124) 29.206 ms 29.482 ms 30.327 ms
6 bbc-technol.edge4.london1.level3.net (126.96.36.199) 29.050 ms 29.930 ms 29.511 ms
7 * * *
8 ae0.er01.telhc.bbc.co.uk (188.8.131.52) 30.094 ms 31.910 ms 30.906 ms
9 184.108.40.206 (220.127.116.11) 29.975 ms 30.208 ms 30.262 ms
10 bbc-vip112.telhc.bbc.co.uk (18.104.22.168) 29.680 ms 29.924 ms 29.223 ms
Now, you'll notice that the latency figures are now consistent, albeit that they're quantitively not brilliant. However, since last weekend, when Vispa first claimed to have fixed the problem, the latencies have been very up-and-down (eg. 29msecs for some hops, 600msecs for other hops). When I told Vispa this on Monday again, they'd by then claimed to have fixed the problem. So, they went away and did something else. Not sure what. The next time I did a traceroute I found that the problem area had moved from the edge.td.vispa.net.uk router to the manchester1.level3.net router, the latter now instead giving a figure of around 500msecs.
I got in touch with Vispa yet again and told them that this was getting beyond a joke; the problem wasn't solved; and they'd been messing around for two weeks now, supposedly fixing it. I told them I was exasperated and made noises to the effect that I'd terminate my account with them if this continued. Low and behold, within an hour the figures had changed and were now consistent (looking pretty much the same as in the traceroute above).
Interestingly, when I ran a traceroute last night, one of the hops - it was actually at london1
.level3 - was now composed of three routers
, each with a latency of around 56 msecs. But, as you can see, that's not now happening. Either Vispa's 'provider' is scurrying around, reallocating resources trying to resolve the problem, or (because of some permanent issue) the routing is automatically adapting and the long latencies are simply being shifted into other routers instead (if you see what I mean).
You being more savvy about routing on the Internet, are you able to conclude anything from this? Do you reckon the provider is BTWholesale? What do you suppose is happening with the long latencies seeming to come and go and to disappear from one router and re-appear instead at another further down the route?