On Fri, Sep 3, 2010 at 7:29 PM, Edward Capriolo <edlinuxg...@gmail.com> wrote: > Interesting, as you may know the web interface code is a dupe of the > cli logic. I know at one point the vector changed from returning > columns to returning rows, or it could have been an oversight on my > part. No unit tests return 1000 rows so no code coverage there. > > I was not aware of this this ticket. I will look at it later tonight. > > On Friday, September 3, 2010, Shrijeet Paliwal <shrij...@rocketfuel.com> > wrote: >> Hi Hive Users, Not sure if closed jiras are attended or not I am submitting >> the comment posicketted here >> (https://issues.apache.org/jira/browse/HIVE-716) to to user list as well: >> >> >> I think we might have bug in HWISessionItem.java. >> I am referring to this change: >> http://svn.apache.org/viewvc/hadoop/hive/trunk/hwi/src/java/org/apache/hadoop/hive/hwi/HWISessionItem.java?r1=817845&r2=817844&pathrev=817845 >> >> Why was this line (343) commented out : res.clear(); >> This is making a query like : " select * from blah limit 1000 " to >> return more that 1000 results (because results are not being flushed) >> -Shrijeet >> >
Shrijeet, Thank you for pointing this out. The ResultBucket that stores the most recent results to view in the interface, shared a the List with the query results. Now the data for the ResultBucket is correctly copied to its own List. Please test out: https://issues.apache.org/jira/browse/HIVE-1615 Edward
