[HADOOP-15790] distcp app should fail if m/r job fails

• Type: Bug
• Status: Open
• Priority: Major
• Resolution: Unresolved
• Affects Version/s: 3.0.0-alpha1, 3.2.0
• Component/s: tools/distcp
• Labels: BB2015-05-TBR
• Tags: distcp distcpv2
• Assignee: Unassigned
• Reporter: David Rosenstrauch
• Created: 30/Sep/13 16:57
• Updated: 10/Sep/21 22:53

Description

I run distcpv2 in a scripted manner. The script checks if the distcp step fails and, if so, aborts the rest of the script. However, I ran into an issue today where the distcp job failed, but my calling script went on its merry way.

Digging into the code a bit more (at https://svn.apache.org/repos/asf/hadoop/common/trunk/hadoop-tools/hadoop-distcp/src/main/java/org/apache/hadoop/tools/DistCp.java), I think I see the issue: the distcp app is not returning an error exit code to the shell when the distcp job fails. This is a big problem, IMO, as it prevents distcp from being successfully used in a scripted environment. IMO, the code should change like so:

Before:

org.apache.hadoop.tools.DistCp.java

//...
  public int run(String[] argv) {
//...
    try {
      execute();
    } catch (InvalidInputException e) {
      LOG.error("Invalid input: ", e);
      return DistCpConstants.INVALID_ARGUMENT;
    } catch (DuplicateFileException e) {
      LOG.error("Duplicate files in input path: ", e);
      return DistCpConstants.DUPLICATE_INPUT;
    } catch (Exception e) {
      LOG.error("Exception encountered ", e);
      return DistCpConstants.UNKNOWN_ERROR;
    }
    return DistCpConstants.SUCCESS;
  }
//...

After:

org.apache.hadoop.tools.DistCp.java

//...
  public int run(String[] argv) {
//...
    Job job = null;
    try {
      job = execute();
    } catch (InvalidInputException e) {
      LOG.error("Invalid input: ", e);
      return DistCpConstants.INVALID_ARGUMENT;
    } catch (DuplicateFileException e) {
      LOG.error("Duplicate files in input path: ", e);
      return DistCpConstants.DUPLICATE_INPUT;
    } catch (Exception e) {
      LOG.error("Exception encountered ", e);
      return DistCpConstants.UNKNOWN_ERROR;
    }

    if (job.isSuccessful()) {
      return DistCpConstants.SUCCESS;
    }
    else {
      return DistCpConstants.UNKNOWN_ERROR;
    }
  }
//...

Issue Links

• is depended upon by: HADOOP-15788 Improve Distcp for long-haul/cloud deployments