Docker环境下JanusGraph无法连接Cassandra后端求助

问题描述 #

自定义构建JanusGraph Docker镜像后，尝试通过Docker Compose让其与Cassandra容器通信，连接失败，报错信息如下：

janusgraph-es  | 2449 [main] ERROR org.apache.tinkerpop.gremlin.server.util.ServerGremlinExecutor  - Could not invoke constructor on class org.janusgraph.graphdb.management.JanusGraphManager (defined by the 'graphManager' setting) with one argument of class Settings
janusgraph-es  | Exception in thread "main" java.lang.RuntimeException: java.lang.reflect.InvocationTargetException
janusgraph-es  |    at org.apache.tinkerpop.gremlin.server.util.ServerGremlinExecutor.<init>(ServerGremlinExecutor.java:91)
janusgraph-es  |    at org.apache.tinkerpop.gremlin.server.GremlinServer.<init>(GremlinServer.java:122)
janusgraph-es  |    at org.apache.tinkerpop.gremlin.server.GremlinServer.<init>(GremlinServer.java:86)
janusgraph-es  |    at org.apache.tinkerpop.gremlin.server.GremlinServer.main(GremlinServer.java:345)
janusgraph-es  | Caused by: java.lang.reflect.InvocationTargetException
janusgraph-es  |    at sun.reflect.NativeConstructorAccessorImpl.newInstance0(Native Method)
janusgraph-es  |    at sun.reflect.NativeConstructorAccessorImpl.newInstance(NativeConstructorAccessorImpl.java:62)
janusgraph-es  |    at sun.reflect.DelegatingConstructorAccessorImpl.newInstance(DelegatingConstructorAccessorImpl.java:45)
janusgraph-es  |    at java.lang.reflect.Constructor.newInstance(Constructor.java:423)
janusgraph-es  |    at org.apache.tinkerpop.gremlin.server.util.ServerGremlinExecutor.<init>(ServerGremlinExecutor.java:80)
janusgraph-es  |    ... 3 more
janusgraph-es  | Caused by: java.lang.IllegalArgumentException: Could not instantiate implementation: org.janusgraph.diskstorage.cql.CQLStoreManager
janusgraph-es  |    at org.janusgraph.util.system.ConfigurationUtil.instantiate(ConfigurationUtil.java:64)
janusgraph-es  |    at org.janusgraph.diskstorage.Backend.getImplementationClass(Backend.java:440)
janusgraph-es  |    at org.janusgraph.diskstorage.Backend.getStorageManager(Backend.java:411)
janusgraph-es  |    at org.janusgraph.graphdb.configuration.builder.GraphDatabaseConfigurationBuilder.build(GraphDatabaseConfigurationBuilder.java:50)
janusgraph-es  |    at org.janusgraph.core.JanusGraphFactory.lambda$open$0(JanusGraphFactory.java:150)
janusgraph-es  |    at org.janusgraph.graphdb.management.JanusGraphManager.openGraph(JanusGraphManager.java:243)
janusgraph-es  |    at org.janusgraph.core.JanusGraphFactory.open(JanusGraphFactory.java:150)
janusgraph-es  |    at org.janusgraph.core.JanusGraphFactory.open(JanusGraphFactory.java:100)
janusgraph-es  |    at org.janusgraph.graphdb.management.JanusGraphManager.lambda$new$0(JanusGraphManager.java:75)
janusgraph-es  |    at java.util.LinkedHashMap.forEach(LinkedHashMap.java:684)
janusgraph-es  |    at org.janusgraph.graphdb.management.JanusGraphManager.<init>(JanusGraphManager.java:74)
janusgraph-es  |    ... 8 more
janusgraph-es  | Caused by: java.lang.reflect.InvocationTargetException
janusgraph-es  |    at sun.reflect.NativeConstructorAccessorImpl.newInstance0(Native Method)
janusgraph-es  |    at sun.reflect.NativeConstructorAccessorImpl.newInstance(NativeConstructorAccessorImpl.java:62)
janusgraph-es  |    at sun.reflect.DelegatingConstructorAccessorImpl.newInstance(DelegatingConstructorAccessorImpl.java:45)
janusgraph-es  |    at java.lang.reflect.Constructor.newInstance(Constructor.java:423)
janusgraph-es  |    at org.janusgraph.util.system.ConfigurationUtil.instantiate(ConfigurationUtil.java:58)
janusgraph-es  |    ... 18 more
janusgraph-es  | Caused by: com.datastax.driver.core.exceptions.NoHostAvailableException: All host(s) tried for query failed (tried: /127.0.0.1:9042 (com.datastax.driver.core.exceptions.TransportException: [/127.0.0.1:9042] Cannot connect))
janusgraph-es  |    at com.datastax.driver.core.ControlConnection.reconnectInternal(ControlConnection.java:268)
janusgraph-es  |    at com.datastax.driver.core.ControlConnection.connect(ControlConnection.java:107)
janusgraph-es  |    at com.datastax.driver.core.Cluster$Manager.negotiateProtocolVersionAndConnect(Cluster.java:1813)
janusgraph-es  |    at com.datastax.driver.core.Cluster$Manager.init(Cluster.java:1726)
janusgraph-es  |    at com.datastax.driver.core.Cluster.init(Cluster.java:214)
janusgraph-es  |    at com.datastax.driver.core.Cluster.connectAsync(Cluster.java:387)
janusgraph-es  |    at com.datastax.driver.core.Cluster.connectAsync(Cluster.java:366)
janusgraph-es  |    at com.datastax.driver.core.Cluster.connect(Cluster.java:311)
janusgraph-es  |    at org.janusgraph.diskstorage.cql.CQLStoreManager.initializeSession(CQLStoreManager.java:320)
janusgraph-es  |    at org.janusgraph.diskstorage.cql.CQLStoreManager.<init>(CQLStoreManager.java:182)
janusgraph-es  |    ... 23 more

提供的配置文件 #

Dockerfile #

FROM openjdk:8-jre-slim-buster

USER root
RUN mkdir -p /janusgraph

COPY . /janusgraph/.

WORKDIR /janusgraph/bin/

CMD ["./gremlin-server.sh"]

Docker Compose #

version: "3.3"
services:
  cassandra:
    image: cassandra:3
    container_name: cassandra
    ports:
      - "9042:9042"
      - "9160:9160"
    healthcheck:
      test: ["CMD-SHELL", "[ $$(nodetool statusgossip) = running ]"]
      interval: 30s
      timeout: 10s
    networks:
      - jce-network
    volumes:
      - app-volume:/var/lib/cassandra
  janusgraph-es:
#    image: docker.io/janusgraph/janusgraph:0.5.3
    build: ./janusgraph-full-0.5.3
    container_name: janusgraph-es
#    environment:
#      JANUS_PROPS_TEMPLATE: cql-es
#      janusgraph.storage.hostname: cassandra
#      janusgraph.index.search.hostname: elastic
#      graph.replace-instance-if-exists: true
    ports:
      - "8182:8182"
    healthcheck:
      test: ["CMD", "bin/gremlin.sh", "-e", "scripts/remote-connect.groovy"]
      interval: 10000s
      timeout: 3000s
      retries: 3
    links:
      - "cassandra"
    networks:
      - jce-network
networks:
  jce-network:
volumes:
  app-volume:

补充说明：已修改JanusGraph配置，指定过Cassandra容器IP和localhost，手动启动JanusGraph可连接Docker中的Cassandra，但将JanusGraph作为Docker服务启动时无法连接。

问题原因与解决方案 #

核心原因 #

从报错NoHostAvailableException: All host(s) tried for query failed (tried: /127.0.0.1:9042)可以明确：JanusGraph容器内部仍在尝试连接本地127.0.0.1的Cassandra服务，而非Docker Compose自定义网络中的Cassandra容器。

具体修复步骤 #

1. 启用JanusGraph环境变量配置
   你的Docker Compose中注释掉了关键的环境变量配置，取消注释并调整：

   environment:
     JANUS_PROPS_TEMPLATE: cql-es
     janusgraph.storage.hostname: cassandra  # Docker Compose服务名，容器间可直接通过此名称访问
     graph.replace-instance-if-exists: true
   

   • janusgraph.storage.hostname设置为Cassandra服务名cassandra，Docker Compose的自定义网络中，容器可通过服务名直接解析并访问对方。

2. 确保自定义镜像支持环境变量覆盖配置
   官方JanusGraph镜像会自动读取环境变量生成对应的配置文件，若自定义镜像无此逻辑，需手动修改配置：

   • 修改/janusgraph/conf/janusgraph.properties，将storage.hostname设置为cassandra
   • 修改/janusgraph/conf/gremlin-server/gremlin-server.yaml，确保加载正确的配置文件

3. 调整启动顺序与健康检查
   JanusGraph启动时Cassandra可能未完全就绪，需配置依赖关联与合理的健康检查：

   • 在JanusGraph服务中添加depends_on，等待Cassandra健康就绪后再启动：

     depends_on:
       cassandra:
         condition: service_healthy
     

   • 调整JanusGraph健康检查参数，缩短间隔确保服务就绪后再标记为健康：

     healthcheck:
       test: ["CMD", "bin/gremlin.sh", "-e", "scripts/remote-connect.groovy"]
       interval: 30s
       timeout: 10s
       retries: 5
     

4. 验证网络连通性
   进入JanusGraph容器，尝试ping Cassandra服务名，确认网络解析正常：

   docker exec -it janusgraph-es ping cassandra
   

完整修改后的Docker Compose示例 #

version: "3.3"
services:
  cassandra:
    image: cassandra:3
    container_name: cassandra
    ports:
      - "9042:9042"
      - "9160:9160"
    healthcheck:
      test: ["CMD-SHELL", "[ $$(nodetool statusgossip) = running ]"]
      interval: 30s
      timeout: 10s
      retries: 3
    networks:
      - jce-network
    volumes:
      - app-volume:/var/lib/cassandra
  janusgraph-es:
    build: ./janusgraph-full-0.5.3
    container_name: janusgraph-es
    environment:
      JANUS_PROPS_TEMPLATE: cql-es
      janusgraph.storage.hostname: cassandra
      graph.replace-instance-if-exists: true
    ports:
      - "8182:8182"
    healthcheck:
      test: ["CMD", "bin/gremlin.sh", "-e", "scripts/remote-connect.groovy"]
      interval: 30s
      timeout: 10s
      retries: 5
    depends_on:
      cassandra:
        condition: service_healthy
    networks:
      - jce-network
networks:
  jce-network:
volumes:
  app-volume:

内容的提问来源于stack exchange，提问作者Hari Krishnan Ramachandran