We're investigating a routing problem that started a few minutes ago affecting broadband and general routing across our network.
This is affecting our services between London and maidenhead, so is affecting some Ethernet circuits too.
Things are getting back to normal now.
Unfortunately, this was caused by human error in a configuration on one of our core switches. The work was being done as part of our investigations to the problems we had last week, and this change was really not meant to cause any issue, except a mistake was made in the configuration. The change was rolled back and the process is being reviewed.