Practice of constructing ten billion relationship knowledge map based on Nebula graph

This article was first published in

Nebula Graph Community  official account

One 、 Project background

Micro LAN is a query technology 、 industry 、 Enterprises 、 Scientific research institutions 、 Application of knowledge atlas of disciplines and their relationships , It contains billions of relationships and billions of entities , In order to make this business run perfectly , Through investigation and research , We use  Nebula Graph  As the main database carrying our knowledge map business , With  Nebula Graph  Product iteration , We finally chose to use it  v2.5.1  Version of  Nebula Graph  As the final version .

Two 、 Why choose  Nebula Graph？

In the field of open source graph database , There are undoubtedly many options , But in order to support the knowledge map service of such large-scale data ,Nebula Graph  Compared with other graph databases, it has the following advantages , This is also our choice  Nebula Graph  Why ：

• Small memory footprint

In our business scenario , our  QPS  Relatively low and without high volatility , At the same time, compared with other graph databases ,Nebula Graph  With less idle time memory consumption , So we can run by using a machine with a lower memory configuration  Nebula Graph  service , This will undoubtedly save us costs .

• Use  multi-raft  Agreement of conformity

multi-raft  Compared to traditional  raft, It not only increases the availability of the system , And the performance is better than traditional  raft  higher . The performance of consensus algorithm mainly depends on whether it allows holes and granularity segmentation , In the application layer, no matter  KV  Database or  SQL , Make good use of these two features , The performance will not be bad . because  raft  The serial commit of is highly dependent on the performance of the state machine , This leads to even in  KV  On , One  key  Of  op  slow , It will significantly slow down others  key. therefore , The key to the performance of a consistent protocol , It must be how the state machine makes it possible to be parallel as much as possible , Even if  multi-raft  The grain size of is relatively coarse （ Compared with  Paxos）, But for those who are not allowed to be empty  raft  Agreement for , There is still a huge improvement .

• The storage side uses  RocksDB  As a storage engine

RocksDB  As a storage engine / embedded database , It has been widely used as the storage end in various databases . More to the point  Nebula Graph  Can be adjusted by  RocksDB  To improve database performance .

• Write fast

Our business needs to write a lot frequently ,Nebula Graph  Even when there is a lot of long text content  vertex  Under the circumstances （ Within cluster 3  Taiwan machine 、3  Copy of the data ,16  Thread insertion ） The insertion speed can also reach  2  ten thousand /s  Speed of insertion , The insertion speed of the edge without attributes can reach  35  ten thousand /s.

3、 ... and 、 Use  Nebula Graph  What problems have we encountered ？

In our knowledge map business , Many scenarios need to show users the paged one degree relationship , At the same time, there are some super nodes in our data , But according to our business scenario ,

The super node must be the node with the highest user access probability , So this can't be simply classified as a long tail problem

; And because we don't have a large number of users , So the cache will not be hit very often , We need a solution to reduce the query latency of users .

give an example ： The business scenario is to query the downstream technologies of this technology , At the same time, we should sort according to the sorting key we set , This sort key is a local sort key . such as , An organization ranks particularly high in a certain field , But in general or in other fields , In this scenario, we must set the sorting attribute to the edge , And the global sorting items are fitted and standardized , So that the variance of data in each dimension is  1, The average is  0, For local sorting , At the same time, it also supports paging operation to facilitate user query .

The statement is as follows ：

MATCH (v1:technology)-[e:technologyLeaf]->(v2:technology) WHERE id(v2) == "foobar" \ 
RETURN id(v1), v1.name, e.sort_value AS sort ORDER BY sort | LIMIT 0,20;


This node has  13  Ten thousand neighbors , In this case, even for  sort_value  Attribute is indexed , The query still takes nearly two seconds . This speed is obviously unacceptable .

We finally chose to use ant financial open source  OceanBase  Database to help us achieve business , The data model is as follows ：

The query statement is as follows ：

SELECT technology.name FROM technology INNER JOIN (SELECT technologydownstream.downstream_id FROM technologydownstream 
WHERE technologydownstream.technology_id = 'foobar' ORDER BY technologydownstream.sort_value DESC LIMIT 0,20) AS t 
WHERE t.downstream_id=technology.id; 


This sentence takes time  80  millisecond . Here is the whole architecture design

Four 、 Use  Nebula Graph  How do we tune ？

I talked about it before.  Nebula Graph  One of the great advantages of is the ability to use native  RocksDB  Parameters for tuning , Reduce the cost of learning , We share the specific meaning of tuning items and some tuning strategies as follows ：

Communication graph database technology ？ Join in  Nebula  Communication group please first

Fill in your  Nebula  Business card

,Nebula  The little assistant will pull you into the group ~~