Sqoop
  1. Sqoop
  2. SQOOP-761

HDFSTextExportExtractor loses lines around partition boundaries

    Details

    • Type: Bug Bug
    • Status: Resolved
    • Priority: Blocker Blocker
    • Resolution: Fixed
    • Affects Version/s: None
    • Fix Version/s: 1.99.1
    • Component/s: None
    • Labels:
      None

      Description

      Blocker for 1.99 release

      1. SQOOP-761-missingdata.txt
        83 kB
        Hari Shreedharan
      2. SQOOP-761.patch
        6 kB
        Hari Shreedharan

        Issue Links

          Activity

          Hide
          Hari Shreedharan added a comment -

          While testing the patch for SQOOP-738, I printed out all lines written out by this class, and the missing lines for the HDFSExtract unit tests.

          Show
          Hari Shreedharan added a comment - While testing the patch for SQOOP-738 , I printed out all lines written out by this class, and the missing lines for the HDFSExtract unit tests.
          Hide
          Hari Shreedharan added a comment -

          The same issue affects the sequence file writer too. Also it seems to affect only the uncompressed data writes.

          Show
          Hari Shreedharan added a comment - The same issue affects the sequence file writer too. Also it seems to affect only the uncompressed data writes.
          Hide
          Hari Shreedharan added a comment - - edited

          Uncommented unit tests. All unit tests pass now. Fixed multiple issues:

          • HdfsTextExportExtractor was using FileSystem.getPos() method - which was causing missing data in uncompressed files - so now explicitly calculate the size.
          • SequenceFile was reading data multiple times, since HdfsSequenceExportExtractor was reading a file till the end of a file, ignoring the end parameter.
          • Added forking to the unit tests, and added more memory.
          • Uncommented the tests in TestHdfsExtract
          Show
          Hari Shreedharan added a comment - - edited Uncommented unit tests. All unit tests pass now. Fixed multiple issues: HdfsTextExportExtractor was using FileSystem.getPos() method - which was causing missing data in uncompressed files - so now explicitly calculate the size. SequenceFile was reading data multiple times, since HdfsSequenceExportExtractor was reading a file till the end of a file, ignoring the end parameter. Added forking to the unit tests, and added more memory. Uncommented the tests in TestHdfsExtract
          Hide
          Jarek Jarcec Cecho added a comment -
          Show
          Jarek Jarcec Cecho added a comment - The patch is in: https://git-wip-us.apache.org/repos/asf?p=sqoop.git;a=commit;h=c41714888808fd56d14b5af11d85b85392a89ff7 Thank you Hari for your hard work! Jarcec

            People

            • Assignee:
              Hari Shreedharan
              Reporter:
              Hari Shreedharan
            • Votes:
              0 Vote for this issue
              Watchers:
              2 Start watching this issue

              Dates

              • Created:
                Updated:
                Resolved:

                Development