At 10:03am this morning, we saw a collection of BGP connections drop in Equinix MA1 (Manchester).
These were spread across LINX MAN1, EquinixIX, and our GTT IP Transit connection.
This caused a couple of minutes of instability whilst BGP reconverged and rerouted traffic automatically.
A few minutes later, we saw these connections establish again.
Based on the varied amount of peers, exchanges, and transit affected we are assuming there must have been a power issue within Equinix MA1 somewhere.
Netnorth’s equipment in Equinix MA1 did not lose power, nor connectivity to other peers.
Equinix have confirmed they saw a power outage within the DC and are investigating the cause.