[
https://issues.apache.org/jira/browse/HIVE-372?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12697705#action_12697705
]
Zheng Shao commented on HIVE-372:
---------------------------------
By just modifying "k=1" to "k=3", the compilation was much faster. We were also
able to identify the correct error like the following:
{code}
hive> create table test (a int b);
FAILED: Parse Error: line 2:25 mismatched input 'b' expecting ) in create
statement
{code}
Originally it was:
{code}
hive> create table test (a int b);
FAILED: Parse Error: line 2:7 mismatched input 'table' expecting TEMPORARY in
create function statement
{code}
I guess the reason is that k=3 makes sure in most cases, we don't need to
backtrack. (We still need backtracking in some cases, since when I switch
backtracking I saw some warnings from antlr).
I will make a patch to change k=1 to k=3, and add a test case for the above,
then I will leave the backtracking problem a bit later.
> Nested UDFs cause _very_ high memory usage when processing query
> ----------------------------------------------------------------
>
> Key: HIVE-372
> URL: https://issues.apache.org/jira/browse/HIVE-372
> Project: Hadoop Hive
> Issue Type: Bug
> Components: Query Processor
> Environment: Fedora Linux, 10x Amazon EC2 (Large Instance w/ 8GB Ram)
> Reporter: Steve Corona
>
> When nesting UDFs, the Hive Query processor takes a large amount of
> time+memory to process the query. For example, I ran something along the
> lines of:
> select trim( trim( trim(trim( trim( trim( trim( trim( trim(column)))))))))
> from test_table;
> This query needs 10GB+ of memory to process before it'll launch the job. The
> amount of memory increases exponentially with each nested UDF.
> Obviously, I am using trim() in this case as a simple example that causes the
> same problem to occur. In my actual use-case I had a bunch of nested
> regexp_replaces.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
