Mailing-List: contact issues-help@drill.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@drill.apache.org
Date: Tue, 29 Nov 2016 01:13:58 +0000 (UTC)
From: "Paul Rogers (JIRA)" <jira@apache.org>
To: issues@drill.apache.org
Message-ID: <JIRA.13023921.1480381437000.384270.1480382038376@Atlassian.JIRA>
In-Reply-To: <JIRA.13023921.1480381437000@Atlassian.JIRA>
References: <JIRA.13023921.1480381437000@Atlassian.JIRA> <JIRA.13023921.1480381437535@arcas>
Subject: [jira] [Comment Edited] (DRILL-5083) IteratorValidator does not
 handle RecordIterator cleanup call to next( )
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit
archived-at: Tue, 29 Nov 2016 01:14:00 -0000


    [ https://issues.apache.org/jira/browse/DRILL-5083?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15703762#comment-15703762 ] 

Paul Rogers edited comment on DRILL-5083 at 11/29/16 1:13 AM:
--------------------------------------------------------------

Stepping further, once the original {{MergeJoinBatch.close( )}} returned, it was called a second time during {{FragmentExeucutor.closeOutResources()}}. The whole bounce up & down occurs a second time.

But, the second time, in {{IteratorValidatorBatchIterator.next()}}, the {{exceptionState}} member is (somehow) no longer set. This lets the operator proceed as if there is no error. It proceeds to call:

{code}
      batchState = incoming.next();
{code}

This eventually hits:

{code}
  public final IterOutcome next(final RecordBatch b) {
    if(!context.shouldContinue()) {
      return IterOutcome.STOP; // <-- Ends up here
    }
    return next(0, b);
  }
{code}

This all seems rather ad-hoc...


was (Author: paul-rogers):
Stepping further, once the original {{MergeJoinBatch.close( )}} returned, it was called a second time during {{FragmentExeucutor.closeOutResources()}}. The whole bounce up & down occurs a second time.

> IteratorValidator does not handle RecordIterator cleanup call to next( )
> ------------------------------------------------------------------------
>
>                 Key: DRILL-5083
>                 URL: https://issues.apache.org/jira/browse/DRILL-5083
>             Project: Apache Drill
>          Issue Type: Bug
>    Affects Versions: 1.8.0
>            Reporter: Paul Rogers
>            Priority: Minor
>
> This one is very confusing...
> In a test with a MergeJoin and external sort, operators are stacked something like this:
> {code}
> Screen
> - MergeJoin
> - - External Sort
> ...
> {code}
> Using the injector to force a OOM in spill, the external sort threw a UserException up the stack. This was handed by:
> {code}
> IteratorValidatorBatchIterator.next( )
> RecordIterator.clearInflightBatches( )
> RecordIterator.close( )
> MergeJoinBatch.close( )
> {code}
> Which does the following:
> {code}
>       // Check whether next() should even have been called in current state.
>       if (null != exceptionState) {
>         throw new IllegalStateException(
> {code}
> But, the exceptionState is set, so we end up throwing an IllegalStateException during cleanup.
> Seems the code should agree: if {{next( )}} will be called during cleanup, then {{next( )}} should gracefully handle that case.


--
This message was sent by Atlassian JIRA
(v6.3.4#6332)