When deploying a new ROFL app, it can happen that the scheduler does not accept a new ROFL app instance due to a lack of capacity, like in the example below.
{"component":"rofl.rofl1...","id":"00000000000000df","level":"info","module":"runtime/scheduler/manager","msg":"evaluating instance","provisioner":"tdx-qemu","runtime_id":"000000000000000000000000000000000000000000000000f80306c9858e7279","runtime_name":"","status":"Created","transfer":false,"ts":"2026-03-21T22:02:00.353614477Z"}
{"component":"rofl.rofl1...","id":"00000000000000df","level":"info","module":"runtime/scheduler/manager","msg":"no more capacity for offer","offer":"0000000000000000","provisioner":"tdx-qemu","runtime_id":"000000000000000000000000000000000000000000000000f80306c9858e7279","runtime_name":"","transfer":false,"ts":"2026-03-21T22:02:00.388221739Z"}The log line indicates we don't have enough local capacity for the chosen offer, but does not let the ROFL provider operator and the user know which resource is lacking - number of instances, vCPUs, memory, or storage.
When deploying a new ROFL app, it can happen that the scheduler does not accept a new ROFL app instance due to a lack of capacity, like in the example below.
The log line indicates we don't have enough local capacity for the chosen offer, but does not let the ROFL provider operator and the user know which resource is lacking - number of instances, vCPUs, memory, or storage.
We can extend the check in https://github.com/oasisprotocol/oasis-sdk/blob/rofl-scheduler/v0.6.3/rofl-scheduler/src/manager.rs#L611 to log more details of what exactly is lacking.