[HADOOP-16360] S3A NullPointerException: null uri host. This can be caused by unencoded / in the password string

• Type: Sub-task
• Status: Resolved
• Priority: Minor
• Resolution: Won't Fix
• Affects Version/s: 3.0.3
• Component/s: fs/s3

Environment

Non AWS store

• Assignee: Unassigned
• Reporter: Ladislav Jech
• Created: 11/Jun/19 10:42
• Updated: 02/Sep/20 14:39
• Resolved: 11/Jun/19 12:18

Description

I am experiencing very old issue appearing now again on Cloudera cluster 6.2. I use following libraries with pyspark job:

• /opt/cloudera/parcels/CDH-6.2.0-1.cdh6.2.0.p0.967373/lib/hadoop/hadoop-common-3.0.0-cdh6.2.0.jar
• /opt/cloudera/parcels/CDH-6.2.0-1.cdh6.2.0.p0.967373/lib/hadoop/hadoop-aws-3.0.0-cdh6.2.0.jar

While trying to write DF to S3 as CSV I get following error:

java.lang.NullPointerException: null uri host. This can be caused by unencoded / in the password string
	at java.util.Objects.requireNonNull(Objects.java:228)
	at org.apache.hadoop.fs.s3native.S3xLoginHelper.buildFSURI(S3xLoginHelper.java:69)
	at org.apache.hadoop.fs.s3a.S3AFileSystem.setUri(S3AFileSystem.java:467)
	at org.apache.hadoop.fs.s3a.S3AFileSystem.initialize(S3AFileSystem.java:234)
	at org.apache.hadoop.fs.FileSystem.createFileSystem(FileSystem.java:3288)
	at org.apache.hadoop.fs.FileSystem.access$200(FileSystem.java:123)
	at org.apache.hadoop.fs.FileSystem$Cache.getInternal(FileSystem.java:3337)
	at org.apache.hadoop.fs.FileSystem$Cache.get(FileSystem.java:3305)
	at org.apache.hadoop.fs.FileSystem.get(FileSystem.java:476)
	at org.apache.hadoop.fs.Path.getFileSystem(Path.java:361)
	at org.apache.spark.sql.execution.datasources.DataSource.planForWritingFileFormat(DataSource.scala:423)
	at org.apache.spark.sql.execution.datasources.DataSource.planForWriting(DataSource.scala:523)
	at org.apache.spark.sql.DataFrameWriter.saveToV1Source(DataFrameWriter.scala:281)
	at org.apache.spark.sql.DataFrameWriter.save(DataFrameWriter.scala:270)
	at org.apache.spark.sql.DataFrameWriter.save(DataFrameWriter.scala:228)
	at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
	at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62)
	at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
	at java.lang.reflect.Method.invoke(Method.java:498)
	at py4j.reflection.MethodInvoker.invoke(MethodInvoker.java:244)
	at py4j.reflection.ReflectionEngine.invoke(ReflectionEngine.java:357)
	at py4j.Gateway.invoke(Gateway.java:282)
	at py4j.commands.AbstractCommand.invokeMethod(AbstractCommand.java:132)
	at py4j.commands.CallCommand.execute(CallCommand.java:79)
	at py4j.GatewayConnection.run(GatewayConnection.java:238)
	at java.lang.Thread.run(Thread.java:748)
// code placeholder

I specify secret key via configuration, not via path (as older bugs reported) and on top of that my Secret key doesn't have any slash, but access key has dash '-' character and 

AWS_HOST_BASE I define with 'http://host.domain.suffix/' form

My code doesn't use secret key in s3 path, but as follows:

sparkSession = SparkSession.builder.getOrCreate() 
sparkContext = sparkSession.sparkContext 
#sparkContext._jsc.hadoopConfiguration().set("fs.s3a.multipart.size", "1000000") 
sparkContext._jsc.hadoopConfiguration().set("fs.s3a.access.key", AWS_ACCESS_KEY_ID) sparkContext._jsc.hadoopConfiguration().set("fs.s3a.secret.key", AWS_SECRET_ACCESS_KEY) sparkContext._jsc.hadoopConfiguration().set("fs.s3a.endpoint", AWS_HOST_BASE) sparkContext._jsc.hadoopConfiguration().set("fs.s3.access.key", AWS_ACCESS_KEY_ID) sparkContext._jsc.hadoopConfiguration().set("fs.s3.secret.key", AWS_SECRET_ACCESS_KEY) sparkContext._jsc.hadoopConfiguration().set("fs.s3.endpoint", AWS_HOST_BASE) sparkContext._jsc.hadoopConfiguration().set("fs.s3n.access.key", AWS_ACCESS_KEY_ID) sparkContext._jsc.hadoopConfiguration().set("fs.s3n.secret.key", AWS_SECRET_ACCESS_KEY) sparkContext._jsc.hadoopConfiguration().set("fs.s3n.endpoint", AWS_HOST_BASE) 

sqlContext = SQLContext(sparkSession.sparkContext) 
# log4j = sparkContext._jvm.org.apache.log4j 
# pylint: disable=W0212

 logger = sparkContext._jvm.org.apache.log4j.LogManager.getLogger("OracleToS3") 
# logger = log4j.LogManager.getlogger(__name__) 
sparkContext.setLogLevel('INFO') 
logger.info("Going to process Oracle tables...") 

for table in Source.table_list:
    logger.info("Reading oracle table into dataframe")
    oracle_table = sparkContext.read \
        .format("jdbc") \
        .option("url", Source.jdbc_string) \
        .option("dbtable", table) \
        .option("user", Source.user) \
        .option("password", Source.password) \
        .option("driver", "oracle.jdbc.driver.OracleDriver") \
        .load()

    # Display schema
    logger.info("Display table schema")
    oracle_table.show()
    logger.info("Display table top 5")
    oracle_table.head(5)
    output_file = "s3a://201906/" + "11/" + table + "_" + time.strftime("%Y%m%d_%H%M%S") +".csv"
    logger.info("Writing table into S3 to file: " + output_file)
    oracle_table\
        .repartition(1)\
        .write \
        .mode("overwrite")\
        .format("csv")\
        .option("header","true") \
        .save("s3a://201906/" + "11/" + table + "_" + time.strftime("%Y%m%d_%H%M%S") +".csv")

Issue Links

• is duplicated by: HADOOP-17241 s3a: bucket names which aren't parseable hostnames unsupported