Wherever a template-reference can occur, it must be possible delete nodes from the template.

[kwatsen]

Requirement

This requirement enables the consumer of a template to remove nodes from a template. For instance, a template may be perfect in many ways, but it sets a value for a leaf when the template-consumer actually wants to let the leaf use its default value.

Non-Requirements

None

[RobertPeschi]

Just for my clarification, is that different form the capability to overrule a template value as per issue #6 ? The case that you raise could be reached by the template consumer explicitly configuring its default value. Or do I miss your point ?

[kwatsen]

Yes, different.

Overriding (per #6) is when, e.g., a template sets a value (e.g., mtu 1500) and the template-consumer overrides it with a more specific value (e.g., mtu 1492). This is the "last one wins" strategy.

This issue/requirement is different where, e.g., the template sets a value (e.g., enabled=true) and the template-consumer deletes it, so now there is no "enabled" node configured causing, e.g., the expanded/flattened configuration to use the YANG-defined 'default' value.

[RobertPeschi]

OK 'get it. Tx.

[derajara]

in case, the node is not deleted by the template-consumer, we will still have 1> value from the template 2> default value from the template-consumer when we try to merge/expand is it not?

[kwatsen]

in case, the node is not deleted by the template-consumer, we will still have 1> value from the template 2> default value from the template-consumer when we try to merge/expand is it not?

I don't understand this question.

[derajara]

I'm looking at a scenario, where the template consumer(s) already are being managed by the same data models, (maybe in a mounted fashion) and if we now want to make these consumers derive from a common template, then we have a situation, where we have a value coming from a template and the existing value from the same data model of each template consumer.

[kwatsen]

I'm looking at a scenario, where the template consumer(s) already are being managed by the same data models, (maybe in a mounted fashion) and if we now want to make these consumers derive from a common template, then we have a situation, where we have a value coming from a template and the existing value from the same data model of each template consumer.

For this scenario, it seems that what’s missing is a step to delete the local configuration redundantly defined by a template consumer. Albeit related, this is a different issue (i.e., delete local vs remote node)

Should we add an issue/requirement for, e.g., action statements to:

• report where (which template) a node’s value comes from
• identify local values having the same value as from a template
• etc.

[derajara]

I think, reporting the source of value will be known with the template-reference, ie , for each template-consumer, we need to provide the reference to a template, that can report where the value comes from. for identifying same values: as per [https://github.com//issues/6] , anyways, we need to provide higher priority to the value coming from individual template-consumer, so even though if it is the same value, it needs to be overwritten