Answer
First, changes and operations are considered separately. Before larger interventions, there must be reliable access, a backup of the current state, a traceable deployment method and, ideally, a test or pre-production environment. Critical background jobs, certificates, domains, storage limits and third-party services are inventoried. Where possible, monitoring and central logging are added.
After that, changes are kept small and released with a fallback plan. Business-critical processes receive real test cases; mere technical availability is not enough. A backup only counts as reliable once it is known what it contains and how a recovery works. For the transition phase, responsibilities between the previous operator, the customer and schukai should be set out in writing. Without this demarcation, an error can easily fall between three parties while nobody is allowed to act.
As of 3 September 2026
Further reading
Related topic: Taking over existing software