We are seeing long boot times at multiple availability zones since 6.45am. Users might experience timeouts when booting new instances.
Core Services is investigating the issue and will provide an update.
This has now been fixed. We've identified the cause as a misbehaving database server. Services have been resumed.