spark sql加载avro

时间:2023-03-09 00:04:15
spark sql加载avro

1、spark sql可以直接加载avro文件,之后再进行一系列的操作,示例:

  

         SparkConf sparkConf = new SparkConf().setAppName("Spark job");
JavaSparkContext javaSparkContext = new JavaSparkContext(sparkConf); SQLContext sqlContext = new SQLContext(javaSparkContext); String FORMAT_CLASS = "com.databricks.spark.avro"; // avro 在hdfs上的路径
String path = "/sqoopdb/pcdas/*.avro";
DataFrame tblarticleautoDf = sqlContext.read().format(FORMAT_CLASS)
.load(path);
tblarticleautoDf.registerTempTable("tableName"); String sql = "select * from tableName";
DataFrame queryDf = sqlContext.sql(sql);
System.out.println(queryDf.count());
System.out.println(queryDf.first());