[ 
https://issues.apache.org/jira/browse/HIVE-1644?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12996227#comment-12996227
 ] 

John Sichi commented on HIVE-1644:
----------------------------------

Stop the press; ignore that last comment.  I hacked something up to use the 
full INSERT compilation, and got that working, but hit the same execution-time 
error as previously.

Then I realized that this is because I was running through with the full test 
case, and there's a problem with the test case itself:  it contains the 
manual-mode steps (including explicitly setting HiveCompactIndexInputFormat), 
and that doesn't play well with the automatic usage.

So then I stripped the test case down to just the automatic usage, and it 
actually seems to be doing something reasonable with the change I mentioned 
previously.

{noformat}
    pctx.getTopOps().putAll(indexQueryPctx.getTopOps());
    pctx.getTopToTable().putAll(indexQueryPctx.getTopToTable());
{noformat}

Try this test case:

{noformat}
CREATE INDEX src_index ON TABLE src(key) as 'COMPACT' WITH DEFERRED REBUILD;
ALTER INDEX src_index ON src REBUILD;

SET hive.input.format=org.apache.hadoop.hive.ql.io.HiveInputFormat;

SELECT * FROM src WHERE key=86 ORDER BY key;

DROP INDEX src_index on src;
{noformat}

If you change the SELECT to EXPLAIN, the plan looks reasonable.  Of course the 
execution won't actually use the index filtering until Jeffery's change is 
rolled in.

So give that a try.


> use filter pushdown for automatically accessing indexes
> -------------------------------------------------------
>
>                 Key: HIVE-1644
>                 URL: https://issues.apache.org/jira/browse/HIVE-1644
>             Project: Hive
>          Issue Type: Improvement
>          Components: Indexing
>    Affects Versions: 0.7.0
>            Reporter: John Sichi
>            Assignee: Russell Melick
>         Attachments: HIVE-1644.1.patch
>
>
> HIVE-1226 provides utilities for analyzing filters which have been pushed 
> down to a table scan.  The next step is to use these for selecting available 
> indexes and generating access plans for those indexes.

-- 
This message is automatically generated by JIRA.
-
For more information on JIRA, see: http://www.atlassian.com/software/jira

        

Reply via email to