I used Jmerter to pressure test the tigerGrap3.2 cluster by adding, deleting, changing and checking points, and found that the throughput of acquiring points was only 2000/s, while neo4j could reach tens of thousands.
Cluster configuration: 128G12 core;
Data set: SF10;
Request sent via restpp, statement: http://192.168.20.60:9000/graph/sf10/vertices/Comment/id.
How can I get more throughput
3 nodes. Neo4j uses the same machine, configuration, and data set.