doc-exports/docs/dli/umn/dli_03_0107.html

How Do I Use Spark to Write Data into a DLI Table?

To use Spark to write data into a DLI table, configure the following parameters:

• fs.obs.access.key
• fs.obs.secret.key
• fs.obs.impl
• fs.obs.endpoint

The following is an example:

import logging
from operator import add
from pyspark import SparkContext

logging.basicConfig(format='%(message)s', level=logging.INFO)

#import local file
test_file_name = "D://test-data_1.txt"
out_file_name = "D://test-data_result_1"

sc = SparkContext("local","wordcount app")
sc._jsc.hadoopConfiguration().set("fs.obs.access.key", "myak")
sc._jsc.hadoopConfiguration().set("fs.obs.secret.key", "mysk")
sc._jsc.hadoopConfiguration().set("fs.obs.impl", "org.apache.hadoop.fs.obs.OBSFileSystem")
sc._jsc.hadoopConfiguration().set("fs.obs.endpoint", "myendpoint")

# red: text_file rdd object
text_file = sc.textFile(test_file_name)

# counts
counts = text_file.flatMap(lambda line: line.split(" ")).map(lambda word: (word, 1)).reduceByKey(lambda a, b: a + b)
# write
counts.saveAsTextFile(out_file_name)

Parent topic: Job Development