Question

我正在尝试在单个查询中创建多个Users节点。它们还在同一查询中与朋友关系链接。当我查询查询时，我看到对于每个属性，我设置了一个关系，进行数据库调用。当每个关系有很多关系和属性时，这会使查询变得昂贵。如何降低此查询的成本？

查询如下：

PROFILE  MERGE (user: User {myId: "1"}) WITH user 
UNWIND [
{myId: "2",userLastName: "2",userAddressId: "1",hasImage: false,randomId: 0},
{myId: "3",userLastName: "3",userAddressId: "3",hasImage: false,randomId: 0},
{myId: "4",userLastName: "4",userAddressId: "1",hasImage: false,randomId: 0},
{myId: "5",userLastName: "5",userAddressId: "1",hasImage: false,randomId: 0},
{myId: "6",userLastName: "6",userAddressId: "1",hasImage: false,randomId: 0},
{myId: "7",userLastName: "7",userAddressId: "1",hasImage: false,randomId: 0},
{myId: "8",userLastName: "8",userAddressId: "1",hasImage: false,randomId: 0},
{myId: "9",userLastName: "9",userAddressId: "1",hasImage: false,randomId: 0} 
] as row  
MERGE (friend: User {myId: row.myId})  ON CREATE SET friend.name = row.userLastName   
MERGE (user)-[c:KNOWS]->(friend)  ON CREATE SET c = row

As you can see in this profile result image there are over 88 db calls in just creating the relation properties from map

如何避免对此查询进行如此多的数据库调用？

注意：已在db上设置CREATE CONSTRAINT ON (user:User) ASSERT user.myId IS UNIQUE。

编辑：

按照this video中的说明，按照@ChristopheWillemsen的说明操作后，我机器上的执行时间从500 + ms下降到约40ms，100到500条记录。

Answer 1

您要求数据库在关系上创建一些属性，因此有一些数据库命中是正常的。

如果你想减少DB的工作量，你可以在ON CREATE SET c = row上显示你想要做的所有SET，如下所示：

MERGE (user)-[c:KNOWS]->(friend)  
 ON CREATE SET c.myId = row.myId,
               c.userLastName=row.userLastName,
               c.userAddressId=row.userAddressId,
               c.hasImage=row.hasImage,
               c.randomId=row.randomId

创建多个属性关系的最佳方式

1 个答案: