UDF in PIG not found

to create new topics or reply. | New User Registration

Tagged: , ,

This topic contains 1 reply, has 2 voices, and was last updated by  Dave 1 year, 7 months ago.

  • Creator
  • #43732

    The error I’m getting is [main] ERROR org.apache.pig.tools.grunt.Grunt - ERROR 2997: Encountered IOException. File access-parser.py does not exist

    access-parser.py is a UDF stored in /tmp/udfs/access-parser.py with user hue and group hdfs, permissions set to 777 (This is the same path, user, and permissions as the default installed udfs I believe)

    The UDF is registered in my script with REGISTER access-parser.py USING jython AS accessParser. I’ve also tried variations with a semicolon at the end of the line, single quotes around the file name, declaring the entire file path, declaring the entire jython path (eg. org.apache.pig.scripting.jython.JythonScriptEngine), completely removing everything from USING onward, etc.

    I’m sure I haven’t accounted for every single variable, but I think I’ve covered the basics in every way I’ve seen a UDF registered, file permissions, and location. I’ve registered python UDF’s in my production environment and using that exact same method doesn’t work here.

    I’m using the default HDP 2.0 Sandbox image in VMWare.

    Does anyone know why Pig can’t find my UDF?

Viewing 1 replies (of 1 total)

You must be to reply to this topic. | Create Account

  • Author
  • #45071


    Hi Michael,

    In your script have you tried registering the UDF by its exact location – /tmp/udfs/access-parser.py ?
    Is this UDF stored on the local filesystem or HDFS ?



Viewing 1 replies (of 1 total)
Hortonworks Data Platform
The Hortonworks Data Platform is a 100% open source distribution of Apache Hadoop that is truly enterprise grade having been built, tested and hardened with enterprise rigor.
Get started with Sandbox
Hortonworks Sandbox is a self-contained virtual machine with Apache Hadoop pre-configured alongside a set of hands-on, step-by-step Hadoop tutorials.
Modern Data Architecture
Tackle the challenges of big data. Hadoop integrates with existing EDW, RDBMS and MPP systems to deliver lower cost, higher capacity infrastructure.