如何配置 Flink 将 Hdfs 用于后端状态和检查点

How to configure Flink to use Hdfs for backend state and checkpoints

我有一个 Flink v1.2、3 个 JobManager、2 个 TaskManager 的设置。我想将 hdfs 用于后端状态和检查点以及 zookeeper storageDir

state.backend: filesystem

state.backend.fs.checkpointdir: hdfs:///[ip:port]/flink-checkpoints
state.checkpoints.dir: hdfs:///[ip:port]/external-checkpoints
high-availability: zookeeper
high-availability.zookeeper.storageDir: hdfs:///[ip:port]/recovery

在我记录的 JobManager 中我有

2017-03-22 17:41:43,559 INFO  org.apache.flink.configuration.GlobalConfiguration            - Loading configuration property: high-availability.zookeeper.client.acl, open
2017-03-22 17:41:43,680 ERROR org.apache.flink.runtime.jobmanager.JobManager                - Error while starting up JobManager
java.io.IOException: The given HDFS file URI (hdfs:///ip:port/recovery/blob) did not describe the HDFS NameNode. The attempt to use a default HDFS configuration, as specified in the 'fs.hdfs.hdfsdefault' or 'fs.hdfs.hdfssite' config parameter failed due to the following problem: Either no default file system was registered, or the provided configuration contains no valid authority component (fs.default.name or fs.defaultFS) describing the (hdfs namenode) host and port.
        at org.apache.flink.runtime.fs.hdfs.HadoopFileSystem.initialize(HadoopFileSystem.java:298)
        at org.apache.flink.core.fs.FileSystem.getUnguardedFileSystem(FileSystem.java:288)
        at org.apache.flink.core.fs.FileSystem.get(FileSystem.java:310)
        at org.apache.flink.runtime.blob.FileSystemBlobStore.<init>(FileSystemBlobStore.java:67)
        at org.apache.flink.runtime.blob.BlobServer.<init>(BlobServer.java:114)
        at org.apache.flink.runtime.jobmanager.JobManager$.createJobManagerComponents(JobManager.scala:2488)
        at org.apache.flink.runtime.jobmanager.JobManager$.startJobManagerActors(JobManager.scala:2643)
        at org.apache.flink.runtime.jobmanager.JobManager$.startJobManagerActors(JobManager.scala:2595)
        at org.apache.flink.runtime.jobmanager.JobManager$.startActorSystemAndJobManagerActors(JobManager.scala:2242)
        at org.apache.flink.runtime.jobmanager.JobManager$.liftedTree3(JobManager.scala:2020)
        at org.apache.flink.runtime.jobmanager.JobManager$.runJobManager(JobManager.scala:2019)
        at org.apache.flink.runtime.jobmanager.JobManager$$anonfun.apply$mcV$sp(JobManager.scala:2098)
        at org.apache.flink.runtime.jobmanager.JobManager$$anonfun.apply(JobManager.scala:2076)
        at org.apache.flink.runtime.jobmanager.JobManager$$anonfun.apply(JobManager.scala:2076)
        at scala.util.Try$.apply(Try.scala:192)
        at org.apache.flink.runtime.jobmanager.JobManager$.retryOnBindException(JobManager.scala:2131)
        at org.apache.flink.runtime.jobmanager.JobManager$.runJobManager(JobManager.scala:2076)
        at org.apache.flink.runtime.jobmanager.JobManager$$anon.call(JobManager.scala:1971)
        at org.apache.flink.runtime.jobmanager.JobManager$$anon.call(JobManager.scala:1969)
        at org.apache.flink.runtime.security.HadoopSecurityContext.run(HadoopSecurityContext.java:43)
        at java.security.AccessController.doPrivileged(Native Method)
        at javax.security.auth.Subject.doAs(Subject.java:422)
        at org.apache.hadoop.security.UserGroupInformation.doAs(UserGroupInformation.java:1548)
        at org.apache.flink.runtime.security.HadoopSecurityContext.runSecured(HadoopSecurityContext.java:40)
        at org.apache.flink.runtime.jobmanager.JobManager$.main(JobManager.scala:1969)
        at org.apache.flink.runtime.jobmanager.JobManager.main(JobManager.scala)
2017-03-22 17:41:43,694 WARN  org.apache.hadoop.security.UserGroupInformation               - PriviledgedActionException as:ubuntu (auth:SIMPLE) cause:java.io.IOException: The given HDFS file URI (hdfs:///ip:port/recovery/blob) did not describe the HDFS NameNode. The attempt to use a default HDFS configuration, as specified in the 'fs.hdfs.hdfsdefault' or 'fs.hdfs.hdfssite' config parameter failed due to the following problem: Either no default file system was registered, or the provided configuration contains no valid authority component (fs.default.name or fs.defaultFS) describing the (hdfs namenode) host and port.
2017-03-22 17:41:43,694 ERROR org.apache.flink.runtime.jobmanager.JobManager                - Failed to run JobManager.
java.io.IOException: The given HDFS file URI (hdfs:///ip:port/recovery/blob) did not describe the HDFS NameNode. The attempt to use a default HDFS configuration, as specified in the 'fs.hdfs.hdfsdefault' or 'fs.hdfs.hdfssite' config parameter failed due to the following problem: Either no default file system was registered, or the provided configuration contains no valid authority component (fs.default.name or fs.defaultFS) describing the (hdfs namenode) host and port.
        at org.apache.flink.runtime.fs.hdfs.HadoopFileSystem.initialize(HadoopFileSystem.java:298)
        at org.apache.flink.core.fs.FileSystem.getUnguardedFileSystem(FileSystem.java:288)
        at org.apache.flink.core.fs.FileSystem.get(FileSystem.java:310)
        at org.apache.flink.runtime.blob.FileSystemBlobStore.<init>(FileSystemBlobStore.java:67)
        at org.apache.flink.runtime.blob.BlobServer.<init>(BlobServer.java:114)
        at org.apache.flink.runtime.jobmanager.JobManager$.createJobManagerComponents(JobManager.scala:2488)
        at org.apache.flink.runtime.jobmanager.JobManager$.startJobManagerActors(JobManager.scala:2643)
        at org.apache.flink.runtime.jobmanager.JobManager$.startJobManagerActors(JobManager.scala:2595)
        at org.apache.flink.runtime.jobmanager.JobManager$.startActorSystemAndJobManagerActors(JobManager.scala:2242)
        at org.apache.flink.runtime.jobmanager.JobManager$.liftedTree3(JobManager.scala:2020)
        at org.apache.flink.runtime.jobmanager.JobManager$.runJobManager(JobManager.scala:2019)
        at org.apache.flink.runtime.jobmanager.JobManager$$anonfun.apply$mcV$sp(JobManager.scala:2098)
        at org.apache.flink.runtime.jobmanager.JobManager$$anonfun.apply(JobManager.scala:2076)
        at org.apache.flink.runtime.jobmanager.JobManager$$anonfun.apply(JobManager.scala:2076)
        at scala.util.Try$.apply(Try.scala:192)
        at org.apache.flink.runtime.jobmanager.JobManager$.retryOnBindException(JobManager.scala:2131)
        at org.apache.flink.runtime.jobmanager.JobManager$.runJobManager(JobManager.scala:2076)
        at org.apache.flink.runtime.jobmanager.JobManager$$anon.call(JobManager.scala:1971)
        at org.apache.flink.runtime.jobmanager.JobManager$$anon.call(JobManager.scala:1969)
        at org.apache.flink.runtime.security.HadoopSecurityContext.run(HadoopSecurityContext.java:43)
        at java.security.AccessController.doPrivileged(Native Method)
        at javax.security.auth.Subject.doAs(Subject.java:422)
        at org.apache.hadoop.security.UserGroupInformation.doAs(UserGroupInformation.java:1548)
        at org.apache.flink.runtime.security.HadoopSecurityContext.runSecured(HadoopSecurityContext.java:40)
        at org.apache.flink.runtime.jobmanager.JobManager$.main(JobManager.scala:1969)
        at org.apache.flink.runtime.jobmanager.JobManager.main(JobManager.scala)
2017-03-22 17:41:43,697 INFO  akka.remote.RemoteActorRefProvider$RemotingTerminator         - Shutting down remote daemon.
2017-03-22 17:41:43,704 INFO  akka.remote.RemoteActorRefProvider$RemotingTerminator         - Remote daemon shut down; proceeding with flushing remote transports.
2

Hadoop 作为单节点集群安装在我在设置中设置的 VM 上。为什么 Flink 要求配置额外的参数? (顺便说一下,它们不在官方文档中)

我认为您必须使用此 URL 模式 hdfs://[ip:port]/flink-checkpoints 来访问符合 hostname:port 规范的 HDFS。

如果您使用的是 Hadoop 配置中的 fs.defaultFS,则无需输入 NameNode 详细信息。