【Spark】pysparkからS3に接続する

Hadoop Configuration

以下のようにcodeまたはcore-site.xmlに記述する。

In code

    sc._jsc.hadoopConfiguration().set("fs.s3.awsAccessKeyId", access_key)
    sc._jsc.hadoopConfiguration().set("fs.s3n.awsAccessKeyId", access_key)
    sc._jsc.hadoopConfiguration().set("fs.s3a.access.key", access_key)
    sc._jsc.hadoopConfiguration().set("fs.s3.awsSecretAccessKey", secret_key)
    sc._jsc.hadoopConfiguration().set("fs.s3n.awsSecretAccessKey", secret_key)
    sc._jsc.hadoopConfiguration().set("fs.s3a.secret.key", secret_key)
    sc._jsc.hadoopConfiguration().set("fs.s3n.impl", "org.apache.hadoop.fs.s3native.NativeS3FileSystem")
    sc._jsc.hadoopConfiguration().set("fs.s3a.impl", "org.apache.hadoop.fs.s3a.S3AFileSystem")
    sc._jsc.hadoopConfiguration().set("fs.s3.impl", "org.apache.hadoop.fs.s3.S3FileSystem")

In core-site.xml

export AWS_ACCESS_KEY_ID=<access-key>
export AWS_SECRET_ACCESS_KEY=<secret-key>

<configuration>
  <property>
    <name>fs.s3n.impl</name>
    <value>org.apache.hadoop.fs.s3native.NativeS3FileSystem</value>
  </property>

  <property>
    <name>fs.s3a.impl</name>
    <value>org.apache.hadoop.fs.s3a.S3AFileSystem</value>
  </property>

  <property>
    <name>fs.s3.impl</name>
    <value>org.apache.hadoop.fs.s3.S3FileSystem</value>
  </property>
</configuration>

実装

import sys
from random import random
from operator import add

from pyspark.sql import SparkSession
from pyspark.conf import SparkConf


if __name__ == "__main__":
    """
        Usage: S3 sample
    """
    access_key = '<access-key>'
    secret_key = '<secret-key>'

    spark = SparkSession\
        .builder\
        .appName("Demo")\
        .getOrCreate()

    sc = spark.sparkContext

    # remove this block if use core-site.xml and env variable
    sc._jsc.hadoopConfiguration().set("fs.s3.awsAccessKeyId", access_key)
    sc._jsc.hadoopConfiguration().set("fs.s3n.awsAccessKeyId", access_key)
    sc._jsc.hadoopConfiguration().set("fs.s3a.access.key", access_key)
    sc._jsc.hadoopConfiguration().set("fs.s3.awsSecretAccessKey", secret_key)
    sc._jsc.hadoopConfiguration().set("fs.s3n.awsSecretAccessKey", secret_key)
    sc._jsc.hadoopConfiguration().set("fs.s3a.secret.key", secret_key)
    sc._jsc.hadoopConfiguration().set("fs.s3n.impl", "org.apache.hadoop.fs.s3native.NativeS3FileSystem")
    sc._jsc.hadoopConfiguration().set("fs.s3a.impl", "org.apache.hadoop.fs.s3a.S3AFileSystem")
    sc._jsc.hadoopConfiguration().set("fs.s3.impl", "org.apache.hadoop.fs.s3.S3FileSystem")

    # fetch from s3, returns RDD
    csv_rdd = spark.sparkContext.textFile("s3n://<bucket-name>/path/to/file.csv")
    c = csv_rdd.count()
    print("~~~~~~~~~~~~~~~~~~~~~count~~~~~~~~~~~~~~~~~~~~~")
    print(c)

    spark.stop()

実行

--packages org.apache.hadoop:hadoop-aws:2.7.1をつけて実行してあげる。

spark-submit --master yarn --deploy-mode client --packages org.apache.hadoop:hadoop-aws:2.7.1 demo.py

参考