Hive / HCatalog Forum

Error from Hive: '403' Message: 'Error while processing statement: null'

  • #37527
    Steen Paulsen
    Participant

    I have created a table in Hive, it consists of four columns:
    – type (string)
    – date (string)
    – entries (int)
    – value (string)

    I can do various SELECT statements directly in Hive via terminal and everything works fine.
    Whenever I try to access the table via the ODBC driver I get the following error:
    “Open Database Connectivity (ODBC) error occurred. state: ‘HY000′. Native Error Code: 35. [Hortonworks][Hardy] (35) Error from Hive: error code: ‘403’ error message: ‘Error while processing statement: null’.”

    I checked the table for NULL values via the following commands:
    – SELECT COUNT(*) FROM tablename WHERE type IS NULL;
    – SELECT COUNT(*) FROM tablename WHERE date IS NULL;
    – SELECT COUNT(*) FROM tablename WHERE entries IS NULL;
    – SELECT COUNT(*) FROM tablename WHERE value IS NULL;
    All of the queries returns “0”.
    To my knowledge this means that there are no NULL values.

    As I can query Hive without any problems via terminal, it leads me to believe that the error is related to the ODBC driver.
    Supporting my case is that the error is thrown immediately after requesting the data, but as the data in the table is 17GB, and the SELECT statement in Hive takes a couple of minutes, this feels like a driver error rather than an error in Hive.

    My setup is three virtual machines in Hyper-V with CentOS 6.4 and HDP 1.3.2 installed via Ambari on a Windows 8 Pro host.
    Any clues on what is going on?

to create new topics or reply. | New User Registration

  • Author
    Replies
  • #38921
    Yi
    Member

    Hi Steen,

    Is the ODBC driver 32bit or 64bit, on Windows 8? Is it configured to connect through hiveserver2?
    When you query with hive client, it does not go through hiveserver2. Can you try query locally using beeline

    beeline> !connect jdbc:hive2://$your-hiveserver-host:port $user org.apache.hive.jdbc.HiveDriver

    and see if you can run the query from beeline?

    if you do lsof -i :$hiveserver2-port, do you see connection from the ODBC host to hiveserver2?

    By default hiveserver2 uses port 10000.

    Also anything in the hiveserver2 log side?

    Thanks,
    Yi

    #39019
    Steen Paulsen
    Participant

    I have tried with both the 32-bit and 64-bit ODBC driver. Both on Windows 8.
    The driver is configured to connect to database “default” through hiveserver2 on port 10000 with hive as user and HIVE as initial catalog.
    I’ve created the ODBC connection via ODBC Data Source Administrator, and the connection confirms as working fine.

    Whenever I try to create the beeline connection as you specified I get the following error:
    # -bash: !connect: event not found

    My exact connection string is:
    beeline> !connect jdbc:hive2://HMaster-Centos:10000 hive org.apache.hive.jdbc.HiveDriver

    The log files are empty

    #39346
    abdelrahman
    Moderator

    Hi Steen,

    What does the log say under /tmp/hive/hive.log? Are you using the latest ODBC driver?

    Thanks
    -Abdelrahman

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

Support from the Experts

A HDP Support Subscription connects you experts with deep experience running Apache Hadoop in production, at-scale on the most demanding workloads.

Enterprise Support »

Become HDP Certified

Real world training designed by the core architects of Hadoop. Scenario-based training courses are available in-classroom or online from anywhere in the world

Training »

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.