conker84 / nodes-2k19

Repository of "Streaming Graph Data with Kafka" talk @ NODES 2k19

Repo of Streaming Graph Data with Kafka talk @ NODES 2k19

Step 1

Execute the docker-compose file via the following command:

docker-compose up -d

You'll find:

• Neo4j Browser @ http://localhost:7474
• Kibana @ http://localhost:5601

Step 2

Create some data by executing the fake data generator:

java -jar kafka-connect-data-generator.jar

This will populate Neo4j (as graph) and Kibana (as indexes)

Step 3

We want to find the most important actors into our graph, and use the result of this computation to sort and reinforce the search operation over Elastic indexes by a hypotetic API

CALL algo.pageRank.stream(
  'MATCH (p:Person) RETURN p.id AS id',
  'MATCH (p1:Person)-->()<--(p2:Person) RETURN distinct p1.id AS source, p2.id AS target',
  {graph:'cypher'}
) YIELD nodeId, score
MATCH (p:Person{id: nodeId})
WHERE p.name is not null
CALL streams.publish('rank', {id: toString(p.id), entity: 'person', properties: {rank: score, name: p.name, born: p.born}})
RETURN nodeId, p.name, score
ORDER BY score DESC