从kafka主题中读取数据并使用spark tempview进行汇总？

Question

i want a read a data from kafka topic, and create spark tempview to group by some columns?

+----+--------------------+
| key|               value|          
+----+--------------------+
|null|{"e":"trade","E":...|
|null|{"e":"trade","E":...|
|null|{"e":"trade","E":...|

but i can't able to aggregate data from tempview?? value column data stored as a String???

Dataset<Row> data = spark
                  .readStream()
                  .format("kafka")
                  .option("kafka.bootstrap.servers", "localhost:9092,localhost:9093")
                  .option("subscribe", "data2-topic")
                  .option("startingOffsets", "latest")
                  .option ("group.id", "test")
                  .option("enable.auto.commit", "true")
                  .option("auto.commit.interval.ms", "1000")          
                  .load();
          data.selectExpr("CAST(key AS STRING)", "CAST(value AS STRING)");
          data.createOrReplaceTempView("Tempdata");
          data.show();
Dataset<Row> df2=spark.sql("SELECT e FROM Tempdata group by e");
df2.show();

Answer 1

value column data stored as a String???

Yes.. Because you CAST(value as STRING)

You'll want to use a from_json function that'll load the row into a proper dataframe that you can search within.

See Databrick's blog on Structured Streaming on Kafka for some examples

If the primary goal is just grouping of some fields, then KSQL might be an alternative.