我正在运行一个概念验证，以连接气流和oracle数据库来存储气流元数据。从文档（https：//airflow.apache.org/docs/apache-airflow/stable/howto/set-up-database.html#database-uri）中，我看到大多数PostgreSQL，MySQL和MSSQL都受到支持，PorsgreSQL是最受欢迎的。然而，由于SQLAlchemy支持Oracle连接，我将airflow docker镜像配置为使用oracle db初始化。
首先，我想澄清的是，我正在Linux Ubuntu 20.4 LTS版本上使用Docker，所有内容都部署在那里。
我将从docker-compose.yml文件开始，它类似于官方airflow文档（https://airflow.apache.org/docs/apache-airflow/2.6.1/docker-compose.yaml）中的文件：

version: '3.8'
x-airflow-common:
  &airflow-common
  #image: ${AIRFLOW_IMAGE_NAME:-apache/airflow:2.6.1-python3.8}
  build: .
  env_file: .env
  user: "${AIRFLOW_UID:-50000}:0"
  environment:
    &airflow-common-env
    AIRFLOW__CORE__EXECUTOR: LocalExecutor
    AIRFLOW__DATABASE__SQL_ALCHEMY_CONN: oracle+cx_oracle://NIKOS:Nikos_123@oracle-db:1521/?service_name=ORCLPDB1
    AIRFLOW__CORE__DAGS_ARE_PAUSED_AT_CREATION: 'true'
    AIRFLOW__CORE__LOAD_EXAMPLES: 'False'
    AIRFLOW__API__AUTH_BACKENDS: 'airflow.api.auth.backend.basic_auth,airflow.api.auth.backend.session'
    AIRFLOW__SCHEDULER__ENABLE_HEALTH_CHECK: 'true'
    _PIP_ADDITIONAL_REQUIREMENTS: ''
  volumes:
    - ${AIRFLOW_PROJ_DIR:-.}/dags:/opt/airflow/dags
    - ${AIRFLOW_PROJ_DIR:-.}/logs:/opt/airflow/logs
    - ${AIRFLOW_PROJ_DIR:-.}/config:/opt/airflow/config
    - ${AIRFLOW_PROJ_DIR:-.}/plugins:/opt/airflow/plugins
  depends_on:
    &airflow-common-depends-on
    oracle-db:
      condition: service_healthy

services:
  oracle-db:
    image: container-registry.oracle.com/database/enterprise:21.3.0.0
    container_name: oracle-db-airflow
    environment:
      ORACLE_SID: ORCLCDB
      ORACLE_PDB: ORCLPDB1
      ORACLE_PWD: Oracle_123
      #ORACLE_EDITION: standard
    ports:
      - 1521:1521
    volumes:
      - oracle-data:/opt/oracle/oradata
      - oracle-backup:/opt/oracle/backup
      - ./create_oracle_user.sql:/create_oracle_user.sql
    healthcheck:
      test: [ "CMD", "/opt/oracle/checkDBStatus.sh"]
      interval: 30s
      timeout: 10s
      retries: 2
    restart: unless-stopped

  airflow-webserver:
    <<: *airflow-common
    command: webserver
    ports:
      - "8080:8080"
    healthcheck:
      test: ["CMD", "curl", "--fail", "http://localhost:8080/health"]
      interval: 30s
      timeout: 10s
      retries: 5
      start_period: 30s
    restart: unless-stopped
    depends_on:
      <<: *airflow-common-depends-on
      airflow-init:
        condition: service_completed_successfully

  airflow-scheduler:
    <<: *airflow-common
    command: scheduler
    healthcheck:
      test: ["CMD", "curl", "--fail", "http://localhost:8974/health"]
      interval: 30s
      timeout: 10s
      retries: 5
      start_period: 30s
    restart: always
    depends_on:
      <<: *airflow-common-depends-on
      airflow-init:
        condition: service_completed_successfully

  airflow-init:
    <<: *airflow-common
    env_file:
      - .env
    user: root
    entrypoint:
      - /bin/bash
      - -c
      - |
        mkdir -p /opt/oracle
        apt-get update
        apt-get install -y unzip
        unzip /opt/instantclient-basic-linux.x64-21.10.0.0.0dbru.zip -d /opt/oracle/
        apt-get install -y libaio1
        sh -c "echo /opt/oracle/instantclient_21_10 > /etc/ld.so.conf.d/oracle-instantclient.conf"
        ldconfig
        ls /opt/oracle/instantclient_21_10 | grep "libclntsh.so"
        /usr/bin/dumb-init -- /entrypoint version
    volumes:
      - ${AIRFLOW_PROJ_DIR:-.}/instantclient-basic-linux.x64-21.10.0.0.0dbru.zip:/opt/instantclient-basic-linux.x64-21.10.0.0.0dbru.zip
    environment:
      <<: *airflow-common-env
      _AIRFLOW_DB_UPGRADE: 'true'
      _AIRFLOW_WWW_USER_CREATE: 'true'
      _AIRFLOW_WWW_USER_USERNAME: ${_AIRFLOW_WWW_USER_USERNAME:-airflow}
      _AIRFLOW_WWW_USER_PASSWORD: ${_AIRFLOW_WWW_USER_PASSWORD:-airflow}
      _PIP_ADDITIONAL_REQUIREMENTS: ''
      LD_LIBRARY_PATH: /opt/oracle/instantclient_21_10

volumes:
  oracle-data:
    external: false
    name: oracla_store
  oracle-backup:
    external: false
    name: oracla_backup

networks:
  default:
    driver: bridge
    name: network_airflow_poc

字符串
oracle-db服务是从container.registry of Oracle获取的官方oracle数据库。数据库运行良好。我可以使用Oracle SQL Developer免费应用程序检查与数据库的连接沿着用户凭据。此外，用户NIKOS也是有效的，并且具有以下权限GRANT CONNECT, CREATE SESSION, CREATE TABLE, CREATE VIEW TO NIKOS;
我的问题是在执行服务airflow-init时发现的。你会注意到，我在docker-compose文件中的这个服务下添加了一些额外的bash命令。这是因为默认情况下未安装cx_Oracle软件包和Oracle Instant Client。因此，我必须执行以下步骤：
步骤1：分别使用以下配置创建Dockerfile和requirements.txt（文档）：

FROM apache/airflow:2.6.1-python3.8
ADD requirements.txt . 
RUN pip install -r requirements.txt

sqlalchemy==1.4.48
cx_Oracle==8.3.0

的数据
并通过取消注解docker service的build参数（build: .）来重建docker airflow镜像。
步骤2：在气流容器内安装Oracle Instant Client。因此，按照官方documentation_1，documentation_2的步骤，我设法克服了以下错误
DPI-1047：找不到64位Oracle客户端库
但是，由于以下日志，整个解决方案无法完成：
x1c 0d1x的数据
完整错误日志here
这个错误基本上说的是表log，我假设是一个airflow metadata表，没有找到。因为从来没有创造过。
奇怪的是，当我使用PostgreSQL数据库作为元数据后端执行整个docker-compose文件并在PGAdmin中检查结果时，我可以清楚地看到初始化的表，如log和其他表。

的
因此，我的问题是如何克服当前在错误日志中遇到的问题？即使Airflow具有有效的SQLAlchemy连接字符串并安装了必要的依赖库，它也无法在oracle db中正确启动其元数据表。