British Telecom

Well, BT finally realized that we’re not going to go away, nor are we going to stop complaining. They sent somebody out to test the line, and to tell them exactly what I’ve been telling them: the problem isn’t between us and the local station, it’s somewhere within the BT network! Our line tested out fine, no problems, and we can always reach the local station. As a matter of fact, we can just about always reach sites within the UK. The problem comes when we’re trying to reach sites which are farther out than the UK, or sometimes for sites within the UK which aren’t cached (see Edge Caching). But will they believe me, when I tell them this? Well, no: I’m just a consumer!

The engineer went through all of his tests, demonstrated that there’s nothing wrong on our end, we can get into the BT network just fine. We just can’t get out!

Running a trace route, we get something like this:

Tracing route to www.sonic.net [209.204.190.64]
over a maximum of 30 hops:

  1    <1 ms    <1 ms    <1 ms  192.168.1.1
  2    28 ms    34 ms    42 ms  217.47.106.58
  3    28 ms    30 ms    27 ms  217.47.106.161
  4     *        *       27 ms  217.32.97.10
  5    27 ms    47 ms    27 ms  217.41.174.21
  6    29 ms    28 ms    28 ms  217.41.174.65
  7    92 ms    28 ms    27 ms  217.41.174.146
  8     *        *       27 ms  217.41.174.50
  9    30 ms    29 ms    29 ms  217.32.97.178
 10     *        *       29 ms  core1-pos14-3.edinburgh.ukcore.bt.net [194.72.17.181]
 11    36 ms    35 ms    44 ms  core1-pos8-0.manchester.ukcore.bt.net [62.6.204.226]
 12    43 ms    42 ms    45 ms  core1-pos0-9-5-0.ealing.ukcore.bt.net [62.172.103.66]
 13    41 ms     *       42 ms  transit1-gig7-0-0.ealing.ukcore.bt.net [62.6.200.106]
 14    41 ms    41 ms    41 ms  t2c1-ge8-0-0.uk-eal.eu.bt.net [166.49.168.17]
 15     *       44 ms    45 ms  t2c2-p3-1.uk-lon1.eu.bt.net [166.49.208.110]
 16    42 ms    42 ms    42 ms  t2a1-ge7-0-0.uk-lon1.eu.bt.net [166.49.135.110]
 17    42 ms    82 ms    42 ms  xe-0-3-0.cr1.lhr1.uk.nlayer.net [195.66.224.37]
 18   123 ms   124 ms   122 ms  xe-2-2-0.cr1.nyc3.us.nlayer.net [69.22.142.9]
 19   216 ms   181 ms   180 ms  xe-1-0-0.cr1.pao1.us.nlayer.net [69.22.142.6]
 20     *        *      179 ms  po1-51.ar1.sfo1.us.nlayer.net [69.22.143.126]
 21   192 ms   193 ms   193 ms  as7065.ge1-1.ar1.sfo1.us.nlayer.net [69.22.153.150]
 22     *        *        *     Request timed out.
 23     *        *        *     Request timed out.
 24     *      183 ms   182 ms  gig0-2.dist2-1.sr.sonic.net [208.201.224.160]
 25     *        *        *     Request timed out.
 26   194 ms   193 ms   195 ms  www.sonic.net [209.204.190.64]

Trace complete.

Along the left is the "hop" - that's our signal, passing through a physical device. In between are the times it took to get from one step to the next. You'll notice that we get lots and lots of timeouts, before even getting a connection to the machine at the other end (in this case, to Sonic.net).

What I find fascinating is that we go through 16 machines before we leave the BT network! Now, sure, we've got to get from here to London, and then out and across the Atlantic. But that particular hand-off doesn't take place until we've been through 17 machines (you'll note that we transition from uk.nlayer.net to us.nlayer.net). There are some further timeouts along the way, of course, but this is by way of aggregating the problem: we have such a shoddy connection within the BT network, that packets are lost. As a matter of fact, 24 hours after clearing our router's logs, we get this:

Transmit Statistics	Receive Statistics
Tx Frames Count	39366	Rx Frames Count	744186
Tx Errors Count	   35	Rx Errors Count	331044
Tx Drops Count	   35	Rx Drops Count	331044

Yes. 331,000 receive frames dropped, out of 744,000 frames attempted. So, the network is losing half of our information! It's not lost, of course - the machines try and try again - but this only adds to the problem. What I really suspect is that BT has attempted to conserve equipment, or hasn't upgraded quickly enough, and that there's just too much of a load on their network. It affects us because we live our online lives primarily in the US - our primary reason for using the network isn't to watch BBC iPlayer, in other words.

We'll keep you posted. It's just one of those frustrations. At least they're listening, this time - we've apparently telephoned them enough to make them realize that we're not going to quit until they fix it.