If a deployment supports in-band task provisioning, then clients should not query /hpke_config with a specified task_id. The reason is:
Aggregators need the first client report for a particular TaskID, in order for the task to be configured.
Clients need the HPKE config to submit reports. If task_id parameter is included when clients query /hpke_config endpoint, the Aggregator may not know that TaskID yet, because it has not received any report for that.
One way is to advise deployment to support a global HPKE config, and also advise clients to not pass task_id to /hpke_config endpoint. The other way is to create the HPKE config on the fly for unseen task ID in /hpke_config.
If a deployment supports in-band task provisioning, then clients should not query
/hpke_config
with a specifiedtask_id
. The reason is:task_id
parameter is included when clients query/hpke_config
endpoint, the Aggregator may not know that TaskID yet, because it has not received any report for that.One way is to advise deployment to support a global HPKE config, and also advise clients to not pass
task_id
to/hpke_config
endpoint. The other way is to create the HPKE config on the fly for unseen task ID in/hpke_config
.