The incident has been deeply analyzed by OGO & the EUW1 Cluster hosting provider engineering teams :
The root cause of the incident was identified : A misconfiguration of the OGO Cluster backend retry strategy led to an erroneous detection of a Syn Flood attack originating from the cluster. The automatic mitigation of this so-called attack by the hosting provider backbone infrastructure led to the blocking of all outgoing traffic from this Cluster.
The Cluster configuration has been corrected and tested so that this scenario won't occur anymore.