重新创建数据库容器时，debezium源任务无法重新连接到postgresql数据库

我们有一个kubernetes集群，其中debezium作为源任务从postgresql运行并写入kafka。德贝齐姆、博士后和Kafka都在各自的豆荚里跑。当postgres pod被删除并且kubernetes重新创建pod时，debezium pod无法重新连接。debezium吊舱的日志：

2018-07-17 08:31:38,311 ERROR  ||  WorkerSourceTask{id=inventory-connector-0} Task is being killed and will not recover until manually restarted   [org.apache.kafka.connect.runtime.WorkerTask]
    2018-07-17 08:31:38,311 INFO   ||  [Producer clientId=producer-4] Closing the Kafka producer with timeoutMillis = 30000 ms.   [org.apache.kafka.clients.producer.KafkaProducer]

debezium继续尝试每隔一段时间刷新未完成的消息，但出现以下异常：

2018-07-17 08:32:38,167 ERROR  ||  WorkerSourceTask{id=inventory-connector-0} Exception thrown while calling task.commit()   [org.apache.kafka.connect.runtime.WorkerSourceTask]
    org.apache.kafka.connect.errors.ConnectException: org.postgresql.util.PSQLException: Database connection failed when writing to copy
    at io.debezium.connector.postgresql.RecordsStreamProducer.commit(RecordsStreamProducer.java:151)
    at io.debezium.connector.postgresql.PostgresConnectorTask.commit(PostgresConnectorTask.java:138)
    at org.apache.kafka.connect.runtime.WorkerSourceTask.commitSourceTask(WorkerSourceTask.java:437)
    at org.apache.kafka.connect.runtime.WorkerSourceTask.commitOffsets(WorkerSourceTask.java:378)
    at org.apache.kafka.connect.runtime.SourceTaskOffsetCommitter.commit(SourceTaskOffsetCommitter.java:108)
    at org.apache.kafka.connect.runtime.SourceTaskOffsetCommitter.access$000(SourceTaskOffsetCommitter.java:45)
    at org.apache.kafka.connect.runtime.SourceTaskOffsetCommitter$1.run(SourceTaskOffsetCommitter.java:82)
    at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511)
    at java.util.concurrent.FutureTask.runAndReset(FutureTask.java:308)
    at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.access$301(ScheduledThreadPoolExecutor.java:180)
    at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.run(ScheduledThreadPoolExecutor.java:294)
    at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
    at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
    at java.lang.Thread.run(Thread.java:748)
    Caused by: org.postgresql.util.PSQLException: Database connection failed when writing to copy
    at org.postgresql.core.v3.QueryExecutorImpl.flushCopy(QueryExecutorImpl.java:942)
    at org.postgresql.core.v3.CopyDualImpl.flushCopy(CopyDualImpl.java:23)
    at org.postgresql.core.v3.replication.V3PGReplicationStream.updateStatusInternal(V3PGReplicationStream.java:176)
    at org.postgresql.core.v3.replication.V3PGReplicationStream.forceUpdateStatus(V3PGReplicationStream.java:99)
    at io.debezium.connector.postgresql.connection.PostgresReplicationConnection$1.doFlushLsn(PostgresReplicationConnection.java:246)
    at io.debezium.connector.postgresql.connection.PostgresReplicationConnection$1.flushLsn(PostgresReplicationConnection.java:239)
    at io.debezium.connector.postgresql.RecordsStreamProducer.commit(RecordsStreamProducer.java:146)
    ... 13 more
    Caused by: java.net.SocketException: Broken pipe (Write failed)
    at java.net.SocketOutputStream.socketWrite0(Native Method)
    at java.net.SocketOutputStream.socketWrite(SocketOutputStream.java:111)
    at java.net.SocketOutputStream.write(SocketOutputStream.java:155)
    at java.io.BufferedOutputStream.flushBuffer(BufferedOutputStream.java:82)
    at java.io.BufferedOutputStream.flush(BufferedOutputStream.java:140)
    at org.postgresql.core.PGStream.flush(PGStream.java:553)
    at org.postgresql.core.v3.QueryExecutorImpl.flushCopy(QueryExecutorImpl.java:939)
    ... 19 more

有没有办法让debezium在上市后重新与postgres建立联系？或者我错过了一些配置？
debezium版本0.8
kubernetes版本1.10.3
postgres 9.6版

看起来这是一个常见的问题，在debezium和kafka中都有开放的特性请求
https://issues.jboss.org/browse/dbz-248
https://issues.apache.org/jira/browse/kafka-5352
虽然这些都是开放的，但这似乎是预期的行为
作为一种解决方法，我将这个活跃度探测器添加到部署中

livenessProbe:
        exec:
          command:
          - sh
          - -ec
          - ipaddress=$(ip addr | grep 'state UP' -A2 | tail -n1 | awk '{print $2}' | cut -f1  -d'/'); reply=$(curl -s $ipaddress:8083/connectors/inventory-connector/status | grep -o RUNNING | wc -l); if [ $reply -lt 2 ]; then exit 1; fi;
        initialDelaySeconds: 30
        periodSeconds: 5

第一个子句获取容器ip地址：

ipaddress=$(ip addr | grep 'state UP' -A2 | tail -n1 | awk '{print $2}' | cut -f1 -d'/');

第二个子句发出请求并统计响应json中“running”的示例：

reply=$(curl -s $ipaddress:8083/connectors/inventory-connector/status | grep -o RUNNING | wc -l);

如果“running”出现少于两次，则第三个子句返回退出代码1

if [ $reply -lt 2 ]; then exit 1; fi

它似乎正在进行初始测试，即重新启动postgresdb会触发debezium容器的重新启动。我猜像这样的脚本（尽管可能是“健壮的”）可以包含在图像中，以便于探测。

重新创建数据库容器时，debezium源任务无法重新连接到postgresql数据库

1条答案

相关问题

热门标签

最新问答