search

5G-MAG / rt-common-shared

Tools common to various projects
Other
3 stars 7 forks source link

Fix M3 interface definitions #14

Closed davidjwbbc closed 1 year ago

davidjwbbc commented 1 year ago

Addresses issues raised in 5G-MAG/rt-common-shared#12.

Fix ContentHostingConfiguration purge request schema to remove extra field. Fix Certificate identifier to be a single value. Add 409 Conflict response to Certificates DELETE interface for when a certificate is still in use by a ContentHostingConfiguration.

davidjwbbc commented 1 year ago

I notice you have added 409 Conflict as a potential error for the destroyServerCertificate operation. Am I right in thinking that this would be returned if an attempt is made to destroy a certificate that is still referenced by a Content Hosting Configuration?

Yes, that's the intention.

rjb1000 commented 1 year ago

@davidjwbbc: Please could you create a new Improvement issue in the 5G-MAG/Standards repository requesting the addition of 409 Conflict as an error for the M1_ServerCertificatesProvisioning_destroyServerCertificate operation so that we can expose this error to the 5GMS Application Provider.

(I think that the issue likely affects all M1 subresources of the Provisioning Session resource, so a more comprehensive title and description is needed.)

(Maybe we could go even wider and overhaul the HTTP response code specifications for service-based operations in all M1 APIs.)

rjb1000 commented 1 year ago

After reviewing the M1 domain model in TS 26.501 figure 5.3.1-1, @davidjwbbc and I concluded that the only possible conflicts today are in attempting to destroy a Server Certificate or Content Preparation Template that is still referenced by a Content Hosting Configuration. And so both _M1_ServerCertificatesProvisioningdestroyServerCertificate and M1_ContentPreparationTemplatesProvisioningdestroyContentPreparationTemplate need to be able to return 409 Conflict. Please document this on a new issue, @davidjwbbc.

For all other subresources of the Provisioning Session, we assume that destroying the Provisioning Session also destroys all subresources. Please verify that this is clearly specified in TS 26.512, @davidjwbbc and raise a second issue if not.

davidjwbbc commented 1 year ago

For all other subresources of the Provisioning Session, we assume that destroying the Provisioning Session also destroys all subresources. Please verify that this is clearly specified in TS 26.512, @davidjwbbc and raise a second issue if not.

This is already stated briefly in TS 26.512 Section 4.3.2.5, but probably ought to be reiterated in Section 7.2.2.

rjb1000 commented 1 year ago

This is already stated briefly in TS 26.512 Section 4.3.2.5, but probably ought to be reiterated in Section 7.2.2.

I think that both clauses could be improved. So please go ahead and raise a new Clarification issue to do this and we can work out the details there.