The cause of the incident has been analysed and the findings were as follows:
In the course of the deployments, a code library element caused an interaction that brought down the mechanism we use to queue up imports of shipments, api creation of shipments and some other activities.
We rolled the change back, corrected the library element and brought the systems back online.
To mitigate against future incidents we’re adding new validations to our tests to catch this in every future deployment.