2026-10-21
Reworked Compute Service Deactivation
Remaining resources in a Compute Service Instance no longer block the deactivation of the Compute Service in the Portal and its API. Previously, you had to remove these resources manually before you could deactivate the Service.
Now the deactivation permanently deletes, among others, the following resources of the Compute Service Instance:
- Servers
- Volumes
- Snapshots
- Images
- Floating IPs
- Routers
- Load Balancers
Deleted resources cannot be recovered. The deactivation dialog in the Project Settings now lists the most important resources that will be deleted.
Because the Kubernetes Service depends on the Compute Service, an active Kubernetes Service in the same Project still
blocks the deactivation, even in the API with force=true.
Deactivate the Kubernetes Service first.
Deactivate a Compute Service via the API
The deactivate endpoint for Compute Services now also accepts requests while resources remain.
If force is missing or false, the API returns 409 Conflict with the list of the most important resources that the
deactivation would delete.
Set force=true to confirm and start the deactivation.
Previously, force=true only succeeded when no such resources were present.
Enhancements and Bug Fixes
- Bug Fix - Compute Service Instance members created without a user: Fixed an issue where a Compute Service Instance member could be created without a user ID if the creation of the backend user failed. As a result, the member list showed duplicate user entries. The Portal now marks such a member as failed and retries transient backend errors instead of dropping them.
- Bug Fix - Server error for unsupported request formats:
Requesting a page in an unsupported format, such as a new session as JSON, now returns
406 Not Acceptableinstead of a500error.