All BT connected ADSL went offline a few minutes ago and are now re-connecting gradually. We are investigating the cause.

Customers may have noticed brief periods of packet loss at 11:22 this morning. Our main transit provider in Maidenhead suffered an outage at that time. Our network detected this within a few seconds, resulting in all traffic automatically being routed via Amsterdam and very little disruption, but apologies to anyone who was affected.

All BT based ADSL lines went down at approximately 08:40 this morning and are gradually re-connecting. Apologies for the inconvenience. If and when we learn from BT the reason, we’ll update this message.

update
BT had a major outage at Stepney Green exchange in London which affected many ISPs fed from there. Lines should have all re-connected now – if any have not please try power-cycling your router.

Some Be connections dropped yesterday evening for a very short period. Some routers have failed to log back in and will need power cycling.

Over the last 24 hours we have had a few slightly less severe repeats of the connectivity problems in ‘mha’. We have identified and disabled a misbehaving virtual server that appears to have been generating traffic that was causing port resets of our FB6000 firewall / edge router ‘grappa’. The FireBrick software team have identified a potential cause for this behaviour and an update to improve robustness will be applied once some bench testing has been carried out. Apologies for the disruption.

Between 13:40 and 14:30 we suffered some co-lo connectivity issues in our ‘mha’ datacentre. We have seen some unexpected entries in the logs of an edge router. All appears well again for now, but we are investigating the significance of these.

There was an unexplained outage just after 02:00 last night, most connections have re established themselves, a few need to have their router re booted.

We are currently working on a few tweaks to our mail server to hopefully prevent re-occurrence of the outages it suffered last week. It will be back soon!

We have just suffered a brief routing flap in our Maidenhead ‘A’ datacentre. The soon to be retired router ‘derek’ went awry and has been reset. Apologies for any strange routes packets may have taken whilst this occurred.

The latest from BT regarding the Reading issue is that a faulty router in Slough has affected all 20CN lines on Reading BRASes and all 21CN lines on Slough BRASes. No estimated time to repair has yet been quoted.