If I understand this correctly, DGraph implemention is more like RDF rather than the property graphs. In that case, the number of nodes will grow exponentially. Is there any numbers/benchmark available to prove the capacity of nodes that DGraph can handle?
Because of the way we store data we don’t consume more space compared to property graph.
Each property is stores as a key value pair.
Example:
we store it as a key value pair with key as “name” and value as Peter. You would need to store predicate and the value in property graph also. Dgraph assigns uid incrementally and our underlying key value store does prefix diffing for keys so there is literally not overhead for storing uid in each key separately.
This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.