java - 组织.apache.spark.SparkException : Task not serializable - JavaSparkContext

Question

我正在尝试运行以下简单的 Spark 代码:

Gson gson = new Gson();
JavaRDD<String> stringRdd = jsc.textFile("src/main/resources/META-INF/data/supplier.json");

JavaRDD<SupplierDTO> rdd = stringRdd.map(new Function<String, SupplierDTO>()
{
    private static final long serialVersionUID = -78238876849074973L;

    @Override
    public SupplierDTO call(String str) throws Exception
    {
        return gson.fromJson(str, SupplierDTO.class);
    }
});

但是在执行 stringRdd.map 语句时抛出以下错误:

org.apache.spark.SparkException: Task not serializable
at org.apache.spark.util.ClosureCleaner$.ensureSerializable(ClosureCleaner.scala:166)
at org.apache.spark.util.ClosureCleaner$.clean(ClosureCleaner.scala:158)
at org.apache.spark.SparkContext.clean(SparkContext.scala:1478)
at org.apache.spark.rdd.RDD.map(RDD.scala:288)
at org.apache.spark.api.java.JavaRDDLike$class.map(JavaRDDLike.scala:78)
at org.apache.spark.api.java.JavaRDD.map(JavaRDD.scala:32)
at com.demo.spark.processor.cassandra.CassandraDataUploader.uploadData(CassandraDataUploader.java:71)
at com.demo.spark.processor.cassandra.CassandraDataUploader.main(CassandraDataUploader.java:47)
Caused by: java.io.NotSerializableException: org.apache.spark.api.java.JavaSparkContext
at java.io.ObjectOutputStream.writeObject0(ObjectOutputStream.java:1184)
at java.io.ObjectOutputStream.defaultWriteFields(ObjectOutputStream.java:1548)
at java.io.ObjectOutputStream.writeSerialData(ObjectOutputStream.java:1509)
at java.io.ObjectOutputStream.writeOrdinaryObject(ObjectOutputStream.java:1432)
at java.io.ObjectOutputStream.writeObject0(ObjectOutputStream.java:1178)
at java.io.ObjectOutputStream.defaultWriteFields(ObjectOutputStream.java:1548)
at java.io.ObjectOutputStream.writeSerialData(ObjectOutputStream.java:1509)
at java.io.ObjectOutputStream.writeOrdinaryObject(ObjectOutputStream.java:1432)
at java.io.ObjectOutputStream.writeObject0(ObjectOutputStream.java:1178)
at java.io.ObjectOutputStream.defaultWriteFields(ObjectOutputStream.java:1548)
at java.io.ObjectOutputStream.writeSerialData(ObjectOutputStream.java:1509)
at java.io.ObjectOutputStream.writeOrdinaryObject(ObjectOutputStream.java:1432)
at java.io.ObjectOutputStream.writeObject0(ObjectOutputStream.java:1178)
at java.io.ObjectOutputStream.writeObject(ObjectOutputStream.java:348)
at org.apache.spark.serializer.JavaSerializationStream.writeObject(JavaSerializer.scala:42)
at org.apache.spark.serializer.JavaSerializerInstance.serialize(JavaSerializer.scala:73)
at org.apache.spark.util.ClosureCleaner$.ensureSerializable(ClosureCleaner.scala:164)
... 7 more

此处“jsc”是我正在使用的 JavaSparkContext 对象。据我所知，JavaSparkContext 不是一个Serializable 对象，不应在将发送给 Spark worker 的任何函数中使用它。

现在，我无法理解的是，JavaSparkContext 的实例是如何发送给工作人员的？我应该在我的代码中更改什么以避免这种情况？

Answer 1

gson 引用将外部类“拉”到闭包的范围内，并带走它的完整对象图。

在这种情况下，在闭包中创建 gson 对象:

public SupplierDTO call(String str) throws Exception {   
   Gson gson = Gson();
   return gson.fromJson(str, SupplierDTO.class);
}

您还可以声明 spark 上下文 transient

如果创建 Gson 实例的成本很高，请考虑使用 mapPartitions 而不是 map。