Hitachi Vantara Pentaho Community Forums
Results 1 to 2 of 2

Thread: Log connection and Log table name auto-insert bad values

  1. #1
    Join Date
    Jun 2015
    Posts
    11

    Default Log connection and Log table name auto-insert bad values

    I'm using the Kettle repository to store my transformations and jobs. In my kettle.properties file I have the definition of my logging variables connection, such as:

    Code:
    KETTLE_JOB_LOG_DB=VM1 - PENTAHO_LOG - DYN
    KETTLE_JOB_LOG_SCHEMA=pentaho_log
    KETTLE_JOB_LOG_TABLE=job_log
    
    KETTLE_TRANS_LOG_DB=VM1 - PENTAHO_LOG - DYN
    KETTLE_TRANS_LOG_SCHEMA=pentaho_log
    KETTLE_TRANS_LOG_TABLE=trans_log
    In the Transformation properties, Logging tab, the fields are left empty since they should search, by default, the variables in my properties file.
    When I close the transformation and open it again, it loads the DB Name directly in the connection and table name, for no reason at all:

    Name:  11-01-2016 5-14-39 PM.jpg
Views: 92
Size:  24.5 KB

    So, clearly, if I try to run the transformation, it'll pop up an error saying it can't find the table VM1 - PENTAHO_LOG - DYN.
    It only adds the connection name to "Log connection" and "Log table name" in the Transformation option, if you go into Step, Performance, Logging channels and Metrics they'll continue empty.

    This also happens with a job, but in this case it will only insert the option in the Log Connection (not in the table name, thus not causing an error when running in the auto-set environment) for the Job log table option.

    Name:  11-01-2016 5-20-14 PM.jpg
Views: 95
Size:  35.1 KB

    I'm not sure if I should report this as a bug, since if it's actually one I believe it would be affecting a lot more people than only myself. Am I doing something wrong?
    Last edited by astro11; 01-12-2016 at 10:40 AM. Reason: fixed grammar

  2. #2
    Join Date
    Jun 2015
    Posts
    11

    Default

    Quick update: by using the variables directly on the field boxes (e.g. ${KETTLE_TRANS_LOG_DB}) everything works fine. But I still think this is a bug, since ideally they should be left blank and Kettle be able to find the parameters in the properties file.
    This would add an enormous amount of work by going into every transformation/job and adding the variables to these fields.

    Also, the tool-tip of these fields suggest that I can leave them blank, e.g.: "The name of the logging connection to use, can also be specified as a variable. You can also specify a default value using variable KETTLE_TRANS_LOG_DB."

    Exploring the bug and the pattern, I found a way to avoid this behavior. I've created a JIRA to see if someone can solve the issue. It's clear to me that this is a bug.
    Last edited by astro11; 01-12-2016 at 02:09 PM. Reason: added JIRA link

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  
Privacy Policy | Legal Notices | Safe Harbor Privacy Policy

Copyright © 2005 - 2019 Hitachi Vantara Corporation. All Rights Reserved.