0

我正在尝试在 neo4j 4.0.4 企业版 CSV 文件中使用 load csv 命令摄取数据，该文件有 15k 行和 9 列，下面是我的 load csv 命令：

USING PERIODIC COMMIT LOAD CSV WITH HEADERS FROM "file:///data/csv_files/NA/admin_2020-05-18_10-19-23_481000/Relationships.csv" 
AS row WITH row WHERE row.`RELATIONSHIP_NAME` IS NOT NULL OR row.`ASSET_1_ID` IS NOT NULL OR row.`ASSET_2_ID` IS NOT NULL 
MERGE (relationship:Relationship{ name:row.`RELATIONSHIP_NAME`, asset2_id:row.`ASSET_2_ID`, asset1_id:row.`ASSET_1_ID` }) 
SET relationship += { name:row.`RELATIONSHIP_NAME`, asset2_id:row.`ASSET_2_ID`, asset1_id:row.`ASSET_1_ID` , relationship_status:row.`RELATIONSHIP_STATUS`, relationship_type_id:row.`RELATIONSHIP_TYPE_ID`, relationship_type:row.`RELATIONSHIP_TYPE`, asset1_type:row.`ASSET_1_TYPE`, asset1_id:row.`ASSET_1_ID`, asset2_type:row.`ASSET_2_TYPE`, asset2_id:row.`ASSET_2_ID`, relationship_comments:row.`RELATIONSHIP_COMMENTS`, modified_at: TIMESTAMP() }

并且没有正在创建的关系，但完成过程大约需要 7 分钟，而且我确实解释了没有急切操作的查询。

dbms.memory.pagecache.size=12100m
dbms.memory.heap.max_size=11900m
dbms.memory.heap.initial_size=11900m

我在这里遗漏了什么还是有办法优化查询。我将如何提高性能？提前致谢

4

1 回答 1

0

您的问题可能也在合并中？一次拥有许多属性。建议合并单个属性，然后添加其他属性。另一件事是在您将要使用的属性上添加索引MERGE。不过，您的查询看起来很奇怪。看起来您正在设置与 MERGE 相同的属性。如果没有任何其他信息，我会将您的查询更改为：

USING PERIODIC COMMIT LOAD CSV WITH HEADERS FROM "file:///data/csv_files/NA/admin_2020-05-18_10-19-23_481000/Relationships.csv" 
AS row WITH row WHERE row.`RELATIONSHIP_NAME` IS NOT NULL OR row.`ASSET_1_ID` IS NOT NULL OR row.`ASSET_2_ID` IS NOT NULL 
CREATE (relationship:Relationship) 
SET relationship += { name:row.`RELATIONSHIP_NAME`, asset2_id:row.`ASSET_2_ID`, asset1_id:row.`ASSET_1_ID` , relationship_status:row.`RELATIONSHIP_STATUS`, relationship_type_id:row.`RELATIONSHIP_TYPE_ID`, relationship_type:row.`RELATIONSHIP_TYPE`, asset1_type:row.`ASSET_1_TYPE`, asset1_id:row.`ASSET_1_ID`, asset2_type:row.`ASSET_2_TYPE`, asset2_id:row.`ASSET_2_ID`, relationship_comments:row.`RELATIONSHIP_COMMENTS`, modified_at: TIMESTAMP() }
于 2020-05-18T12:55:30.867 回答