sqllineage解析sql列级血缘并提交到datahub

版本信息

python 3.8.16

datahub v0.10.0

操作记录

安装datahub v0.10.0

详见datahub官网 A Metadata Platform for the Modern Data Stack | DataHub

执行命令

python3 -m pip install --upgrade pip wheel setuptools

python3 -m pip install --upgrade acryl-datahub==0.10.0

查看版本

python3 -m datahub version

datahub 快速部署

将datahub v0.10.0分支下的docker-compose-without-neo4j.quickstart.yml文件准备到本地

datahub/docker-compose-without-neo4j.quickstart.yml at v0.10.0 · datahub-project/datahub · GitHub

确保以下端口未被占用

• 3306 for MySQL
• 9200 for Elasticsearch
• 9092 for the Kafka broker
• 8081 for Schema Registry
• 2181 for ZooKeeper
• 9002 for the DataHub Web Application (datahub-frontend)
• 8080 for the DataHub Metadata Service (datahub-gms)

如有占用在命令行传参进行替换

datahub docker quickstart --mysql-port 53306

执行

python3 -m datahub docker quickstart -f ./docker-compose-without-neo4j.quickstart.yml --version v0.10.0

开始拉取镜像

成功构建容器，datahub启动成功

访问hadoop105:9002

输入账号、密码datahub

元数据摄取

安装hive插件

python3 -m pip install 'acryl-datahub[hive]'

安装过程中报错

尝试安装依赖项

yum -y install gcc gcc-c++ python-devel.x86_64 cyrus-sasl-devel.x86_64 gcc-c++.x86_64

再次安装hive插件

检查datahub插件

python3 -m datahub check plugins

hive插件成功安装

编写摄取hive元数据的配置文件

source:
  type: "hive"
  config: 
    host_port: "hadoop102:10000" # hiveserver2 

sink:
  type: "datahub-rest"
  config:
    server: "http://hadoop105:8080" # datahub gms server

开始摄取hive元数据

python3 -m datahub ingest -c ./hive-metadata-ingestion.yml

元数据摄取完成

进入web页面查看

通过sqllineage获取指定sql文件中HiveSQL的字段级血缘关系，并将结果提交到datahub

参考datahub官方文档给出的提交细粒度血缘的脚本datahub/lineage_emitter_dataset_finegrained.py at master · datahub-project/datahub · GitHub

参考sqllineage文档Getting Started — sqllineage 1.3.7 documentation

结合sqllineage，获取指定sql的列级血缘，再调用datahub rest api，将结果提交到datahub

具体py代码如下

from sqllineage.runner import LineageRunner

import datahub.emitter.mce_builder as builder
from datahub.emitter.mcp import MetadataChangeProposalWrapper
from datahub.emitter.rest_emitter import DatahubRestEmitter
from datahub.metadata.com.linkedin.pegasus2avro.dataset import (
    DatasetLineageType,
    FineGrainedLineage,
    FineGrainedLineageDownstreamType,
    FineGrainedLineageUpstreamType,
    Upstream,
    UpstreamLineage,
)
import sys

'''
    解析目标sql文件的HiveSQL生成列级血缘，提交到datahub
    sql文件路径作为命令行参数传入脚本
    提交到datahub的platform = hive
'''

# 库名设置
def datasetUrn(tableName):
    return builder.make_dataset_urn("hive", tableName)  # platform = hive

# 表、列级信息设置
def fieldUrn(tableName, fieldName):
    return builder.make_schema_field_urn(datasetUrn(tableName), fieldName)

# 目标sql文件路径
sqlFilePath = sys.argv[1]

sqlFile = open(sqlFilePath, mode='r', encoding='utf-8')

sql = sqlFile.read().__str__()

# 获取sql血缘
result = LineageRunner(sql)

# 获取sql中的下游表名
targetTableName = result.target_tables[0].__str__()

print(result)

print('===============')

# 打印列级血缘结果
result.print_column_lineage()

print('===============')

# 获取列级血缘
lineage = result.get_column_lineage

# 字段级血缘list
fineGrainedLineageList = []

# 用于冲突检查的上游list
upStreamsList = []

# 遍历列级血缘
for columnTuples in lineage():
    # 上游list
    upStreamStrList = []

    # 下游list
    downStreamStrList = []

    # 逐个字段遍历
    for column in columnTuples:

        # 元组中最后一个元素为下游表名与字段名，其他元素为上游表名与字段名

        # 遍历到最后一个元素，为下游表名与字段名
        if columnTuples.index(column) == len(columnTuples) - 1:
            downStreamFieldName = column.raw_name.__str__()
            downStreamTableName = column.__str__().replace('.' + downStreamFieldName, '').__str__()

            # print('下游表名：' + downStreamTableName)
            # print('下游字段名：' + downStreamFieldName)

            downStreamStrList.append(fieldUrn(downStreamTableName, downStreamFieldName))
        else:
            upStreamFieldName = column.raw_name.__str__()
            upStreamTableName = column.__str__().replace('.' + upStreamFieldName, '').__str__()

            # print('上游表名：' + upStreamTableName)
            # print('上游字段名：' + upStreamFieldName)

            upStreamStrList.append(fieldUrn(upStreamTableName, upStreamFieldName))

            # 用于检查上游血缘是否冲突
            upStreamsList.append(Upstream(dataset=datasetUrn(upStreamTableName), type=DatasetLineageType.TRANSFORMED))

    fineGrainedLineage = FineGrainedLineage(upstreamType=FineGrainedLineageUpstreamType.DATASET,
                                            upstreams=upStreamStrList,
                                            downstreamType=FineGrainedLineageDownstreamType.FIELD_SET,
                                            downstreams=downStreamStrList)

    fineGrainedLineageList.append(fineGrainedLineage)

fieldLineages = UpstreamLineage(
    upstreams=upStreamsList, fineGrainedLineages=fineGrainedLineageList
)

lineageMcp = MetadataChangeProposalWrapper(
    entityUrn=datasetUrn(targetTableName),  # 下游表名
    aspect=fieldLineages
)

# 调用datahub REST API
emitter = DatahubRestEmitter('http://datahub-gms:8080') # datahub gms server

# Emit metadata!
emitter.emit_mcp(lineageMcp)

执行py脚本

python3 sql-lineage-to-datahub.py target.sql

查看web界面