Identifying relationships are not determined by an existence dependency between two entity sets. An existence dependency generally means an entity set must participate totally in a relationship.
Identifying relationships mean one entity set depends on another for identity. The dependent entity set is called a weak entity set, since it can't be identified by its own attributes alone.
In your example, I assume oid
is an auto-incremented column. That makes it a surrogate key which identifies Order
, and Order
is then a regular entity set since it can be identified by its own attributes. This means the relationship between the two entity sets isn't identifying.
The relationship is represented by the pair of columns (oid, cid)
. oid
should be PK, so that each order (oid
) can only be associated with one customer (cid
). However, there's no unique constraint on cid
, which means each cid
can be associated with multiple oid
. Hence, it's a one-to-many relationship.
Also see my answer to the following question: is optionality (mandatory, optional) and participation (total, partial) are same?