This incident occurred after a deployment made early in the morning. The release content is not the cause of the incident.
The problem lies in restarting the services affected by the deployment. This action requires a large amount of memory and CPU. It seems that the node used this morning was too limited in CPU preventing the bot services from starting properly.
Around 9:25am (CEST), we manually ejected this node in order to force a reboot on a node with enough CPU.
We will adapt the server configuration to prevent this problem in the future.