针对kubernetes的Spark - Azure Blob存储凭据问题

Question

当应用程序jar存储在Azure Blob存储容器中时,我尝试为kubernetes运行Spark时出现以下问题:

2018-10-18 08:48:54 INFO  DAGScheduler:54 - Job 0 failed: reduce at SparkPi.scala:38, took 1.743177 s
Exception in thread "main" org.apache.spark.SparkException: Job aborted due to stage failure: Task 0 in stage 0.0 failed 4 times, most recent failure: Lost task 0.3 in stage 0.0 (TID 6, 10.244.1.11, executor 2): org.apache.hadoop.fs.azure.AzureException: org.apache.hadoop.fs.azure.AzureException: No credentials found for account datasets83d858296fd0c49b.blob.core.windows.net in the configuration, and its container datasets is not accessible using anonymous credentials. Please check if the container exists first. If it is not publicly available, you have to provide account credentials.
    at org.apache.hadoop.fs.azure.AzureNativeFileSystemStore.createAzureStorageSession(AzureNativeFileSystemStore.java:1086)
    at org.apache.hadoop.fs.azure.AzureNativeFileSystemStore.initialize(AzureNativeFileSystemStore.java:538)
    at org.apache.hadoop.fs.azure.NativeAzureFileSystem.initialize(NativeAzureFileSystem.java:1366)
    at org.apache.hadoop.fs.FileSystem.createFileSystem(FileSystem.java:3242)
    at org.apache.hadoop.fs.FileSystem.access$200(FileSystem.java:121)
    at org.apache.hadoop.fs.FileSystem$Cache.getInternal(FileSystem.java:3291)
    at org.apache.hadoop.fs.FileSystem$Cache.get(FileSystem.java:3259)
    at org.apache.hadoop.fs.FileSystem.get(FileSystem.java:470)
    at org.apache.spark.util.Utils$.getHadoopFileSystem(Utils.scala:1897)
    at org.apache.spark.util.Utils$.doFetchFile(Utils.scala:694)
    at org.apache.spark.util.Utils$.fetchFile(Utils.scala:476)
    at org.apache.spark.executor.Executor$$anonfun$org$apache$spark$executor$Executor$$updateDependencies$5.apply(Executor.scala:755)
    at org.apache.spark.executor.Executor$$anonfun$org$apache$spark$executor$Executor$$updateDependencies$5.apply(Executor.scala:747)
    at scala.collection.TraversableLike$WithFilter$$anonfun$foreach$1.apply(TraversableLike.scala:733)
    at scala.collection.mutable.HashMap$$anonfun$foreach$1.apply(HashMap.scala:99)
    at scala.collection.mutable.HashMap$$anonfun$foreach$1.apply(HashMap.scala:99)
    at scala.collection.mutable.HashTable$class.foreachEntry(HashTable.scala:230)
    at scala.collection.mutable.HashMap.foreachEntry(HashMap.scala:40)
    at scala.collection.mutable.HashMap.foreach(HashMap.scala:99)
    at scala.collection.TraversableLike$WithFilter.foreach(TraversableLike.scala:732)
    at org.apache.spark.executor.Executor.org$apache$spark$executor$Executor$$updateDependencies(Executor.scala:747)
    at org.apache.spark.executor.Executor$TaskRunner.run(Executor.scala:312)
    at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
    at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
    at java.lang.Thread.run(Thread.java:748)
Caused by: org.apache.hadoop.fs.azure.AzureException: No credentials found for account datasets83d858296fd0c49b.blob.core.windows.net in the configuration, and its container datasets is not accessible using anonymous credentials. Please check if the container exists first. If it is not publicly available, you have to provide account credentials.
    at org.apache.hadoop.fs.azure.AzureNativeFileSystemStore.connectUsingAnonymousCredentials(AzureNativeFileSystemStore.java:863)
    at org.apache.hadoop.fs.azure.AzureNativeFileSystemStore.createAzureStorageSession(AzureNativeFileSystemStore.java:1081)
    ... 24 more

我用来启动作业的命令是:

/opt/spark/bin/spark-submit
    --master k8s://<my-k8s-master>
    --deploy-mode cluster
    --name spark-pi
    --class org.apache.spark.examples.SparkPi
    --conf spark.executor.instances=5
    --conf spark.kubernetes.container.image=<my-image-built-with-wasb>
    --conf spark.kubernetes.namespace=<my-namespace>
    --conf spark.kubernetes.authenticate.driver.serviceAccountName=spark
    --conf spark.kubernetes.driver.secrets.spark=/opt/spark/conf
    --conf spark.kubernetes.executor.secrets.spark=/opt/spark/conf
wasb://<my-container-name>@<my-account-name>.blob.core.windows.net/spark-examples_2.11-2.3.2.jar 10000

我有一个以spark下列内容命名的k8s秘密:

apiVersion: v1
kind: Secret
metadata:
  name: spark
  labels:
    app: spark
    stack: service
type: Opaque
data:
  core-site.xml: |-
    {% filter b64encode %}
    <configuration>
        <property>
            <name>fs.azure.account.key.<my-account-name>.blob.core.windows.net</name>
            <value><my-account-key></value>
        </property>
        <property>
            <name>fs.AbstractFileSystem.wasb.Impl</name>
            <value>org.apache.hadoop.fs.azure.Wasb</value>
        </property>
    </configuration>
    {% endfilter %}

驱动程序pod设法下载存储在Azure Blob存储中的容器中的jar依赖项.从这个日志片段中可以看出:

2018-10-18 08:48:16 INFO  Utils:54 - Fetching wasb://<my-container-name>@<my-account-name>.blob.core.windows.net/spark-examples_2.11-2.3.2.jar to /var/spark-data/spark-jars/fetchFileTemp8575879929413871510.tmp
2018-10-18 08:48:16 INFO  SparkPodInitContainer:54 - Finished downloading application dependencies.

如何让执行程序窗格获取存储在core-site.xml从k8s密钥挂载的文件中的凭据？我错过了什么？

Answer 1

我通过将以下配置添加到spark-submit来解决它

--conf spark.hadoop.fs.AbstractFileSystem.wasb.Impl=org.apache.hadoop.fs.azure.Wasb
--conf spark.hadoop.fs.azure.account.key.${STORAGE_ACCOUNT_NAME}.blob.core.windows.net=${STORAGE_ACCOUNT_KEY}