CODEWITHSUNDEEP
optimizationneo4jgraph-databases
Cassie
Cassie

Reputation: 3099

Match query optimization for multiple nodes Neo4j

I am running a query on 3 tables which are connected with the same id of an application. However, this query runs too long to get any result. How can I optimize this query? The application_id by which both edges are created should be the same for both Sms and Pcb nodes.

This is the query itself:

 MATCH (pcb:PhoneContactsBook)-[:APP_PCB]->(a:Applications)-[:APP_SMS]->(sms:Sms)
    RETURN distinct a.application_id

To get the query plan, I run this query with LIMIT 200,000 and I get 1046648 total db hits in 1251 ms.

Here is the query plan:

enter image description here

Upvotes: 0

Views: 36

Answers (1)

logisima
logisima

Reputation: 7478

Firstly, to get the query plan, you can use EXPLAIN instead of PROFILE :

  • EXPLAIN do not run the query and only give you the query plan of the query
  • PROFILE executes the query, and give you the result of your query + its query plan

I don't see any optimization here, because you are asking the database to give you all the instance of a specific pattern and there is no where clause, no sub-statement, no aggregation ...

But it seems that you have a lot instances of this pattern in your database (more than 200 000), that's why this query takes some times (~6 microsecondes per pattern is not bad ^^)

Upvotes: 1

Related Questions