Uploaded image for project: 'IMPALA'
  1. IMPALA
  2. IMPALA-5993

Fix the file offset in value parsing error messages when scanning text files

    XMLWordPrintableJSON

Details

    • Bug
    • Status: Open
    • Major
    • Resolution: Unresolved
    • Impala 2.10.0, Impala 2.11.0
    • None
    • Backend
    • ghx-label-3

    Description

      When the text scanner hits an error, it always prints the end of the file as the offset, even if the error occurs in the middle of the file. For example:

      WARNINGS: Error converting column: 3 to INT
Error parsing row: file: hdfs://localhost:20500/home/lv/sample_07/sample_07, before offset: 46055
Error converting column: 3 to INT

      We should instead print the offset of the problematic value to make debugging easier.

      Attachments

        1. sample_07
          45 kB
          Lars Volker

        Issue Links

          is related to

          Improvement - An improvement or enhancement to an existing feature or task. IMPALA-6971 Log flooding with unhelpful parsing error message

          • Minor - Minor loss of function, or other problem where easy workaround is present.
          • Open
          relates to

          Bug - A problem which impairs or prevents the functions of the product. IMPALA-3859 HdfsScanNode::LogRowParseError() will log table data

          • Critical - Crashes, loss of data, severe memory leak.
          • Resolved

          Activity

            People

              Unassigned Unassigned
              lv Lars Volker
              Votes:
              1 Vote for this issue
              Watchers:
              5 Start watching this issue

              Dates

                Created:
                Updated: