Apoc.merge.node VS merge query

Neo4j Graph Platform Procedures & APOC
, , ,
1

Hi, Suppose i have import millions of data in my graph. so basically i have two option to create the nodes in graph.
First way is using simple Merge query as written below
MERGE (p:Landing{IP:row.IP,NUM:row.NUM})
ON CREATE SET
p.IP_Null='N',
p.REL_null='N',
p.NUM=row.NUM,
p.IP=row.IP
ON MATCH SET
p.IP_Null='N',
p.REL_null='N',
p.NUM=row.NUM,
p.IP=row.IP

return p

Second one is using apoc.,merge.node procedure.
call apoc.Merge.node ([p:Landing],{IP:row.IP,NUM:row.NUM},{
p.IP_Null='N',
p.REL_null='N',
p.NUM=row.NUM,
p.IP=row.IP},{p.IP_Null='N',
p.REL_null='N',
p.NUM=row.NUM,
p.IP=row.IP} yield node
return node

My doubt is how we can understand which query is much faster.
Note i am trying this query in neo4j V3.5.

2

I doubt these two would be noticeably different. I think the benefit of the apoc approach is the flexibility with setting the label dynamically.

Just a note, your label list in the apoc call should be [“Landing”], not [p:Landing]. Also, the property maps for creating and matching the node are neo4j maps. You have syntax similar to SET operations. Remove the reference to ‘p’ and change the maps to key value pairs.

Also, your biggest performance will be gained with a composite index on the Landing label and the two properties IP and NUM.