drill-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Sudheesh Katkam" <skat...@maprtech.com>
Subject Re: Review Request 34191: DRILL-3052: Canceling a fragment executor before it starts running will cause the Foreman to wait indefinitely for a terminal message from that fragment
Date Thu, 14 May 2015 21:03:50 GMT

This is an automatically generated e-mail. To reply, visit:

(Updated May 14, 2015, 9:03 p.m.)

Review request for drill, Jacques Nadeau and Venki Korukanti.


Addressed review changes, and rebased on current master [583ca4a]

Repository: drill-git


[DRILL-3052](https://issues.apache.org/jira/browse/DRILL-3052), [DRILL-3066](https://issues.apache.org/jira/browse/DRILL-3066):
FragmentExecutor must cleanup exactly once
Cleanup includes, in order:
1) closing out resources,
2) updating to the correct terminal state, and
3) sending the state to QueryManager exactly once

In DRILL-3053 scenario, sendFinalState() is never called
In DRILL-3066 scenario, closeOutResources() is called twice

Diffs (updated)

  exec/java-exec/src/main/java/org/apache/drill/exec/work/fragment/FragmentExecutor.java 8c49d68

  exec/java-exec/src/test/java/org/apache/drill/exec/server/TestDrillbitResilience.java f95fbe1

Diff: https://reviews.apache.org/r/34191/diff/


Passes all unit test and regression tests. Started another build.


Sudheesh Katkam

  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message