Details
-
Type:
Bug
-
Status: Resolved
-
Priority:
Major
-
Resolution: Duplicate
-
Affects Version/s: 3.11.0
-
Fix Version/s: None
-
Component/s: con.spark
-
Labels:None
-
Epic Link:
Description
I have a simple pySpark snippet:
subprocess.call(["hadoop", "fs", "-rm", "-R", "mnist_model"])
Executing the snippet often returns the following exception:
com.fasterxml.jackson.core.JsonParseException: Unrecognized token 'Moved': was expecting ('true', 'false' or 'null') at [Source: Moved: 'hdfs://namenode:port/user/username/mnist_model' to trash at: hdfs://namenode:port/user/username/.Trash/Current; line: 1, column: 6]
I replaced the namenode host and port with namenode:port above, and the username with "username".
Can Hue not handle certain chars in the snippet output? The underlying "hadoop fs" command should return the following, and it seems like the ":" or single quote char is tripping it up.
Moved: 'hdfs://namenode:port/user/username/mnist_model' to trash at: hdfs://namenode:port/user/username/.Trash/Current