A common case of an address pool is something like "10.0.0.2 - 10.0.0.99".
This cannot be done because the pool-prefix is mandatory. I propose to make it optional and to explain that it SHOULD be used in the case the pool range can be expressed as a prefix. Note the code to convert prefix to range and the opposite is both easy to write and to find.
Note that max-address-count is a threshold which has a "disabled" value so there is no problem to leave it mandatory. Same for max-pd-space-utilization in pd pools.
Last detail: it is not clear if the pool-id scope is the network range or is global. I suggest to add a statement explaining it has a network range scope. Same for pd pools which have their own network range scope. Of course this is not true for network range ids which must be global.
A common case of an address pool is something like "10.0.0.2 - 10.0.0.99".
This cannot be done because the pool-prefix is mandatory. I propose to make it optional and to explain that it SHOULD be used in the case the pool range can be expressed as a prefix. Note the code to convert prefix to range and the opposite is both easy to write and to find.
Note that max-address-count is a threshold which has a "disabled" value so there is no problem to leave it mandatory. Same for max-pd-space-utilization in pd pools.
Last detail: it is not clear if the pool-id scope is the network range or is global. I suggest to add a statement explaining it has a network range scope. Same for pd pools which have their own network range scope. Of course this is not true for network range ids which must be global.