Re: Full text query in Phoenix

2016-09-19 Thread Cheyenne Forbes
Hi James, Thanks a lot, I found a link showing how to integrate hbase with lucene https://itpeernetwork.intel.com/idh-hbase-lucene-integration/

Re: Using COUNT() with columns that don't use COUNT() when the table is join fails

2016-09-19 Thread Steve Terrell
I'm not an expert in traditional SQL or in Phoenix SQL, but my best guess is "probably not". But I'm curious as to why you would like to avoid the group by or the list of columns. I know it looks very wordy, but are there any technical reasons? In my experience SQL is hard to read by human eyes

Re: Using COUNT() with columns that don't use COUNT() when the table is join fails

2016-09-19 Thread Cheyenne Forbes
I was wondering because it seems extra wordy

Re: Using COUNT() with columns that don't use COUNT() when the table is join fails

2016-09-19 Thread Steve Terrell
Hi! I think you need something like group by u.first_name on the end. Best guess. :) On Sun, Sep 18, 2016 at 11:03 PM, Cheyenne Forbes < cheyenne.osanu.for...@gmail.com> wrote: > this query fails: > > SELECT COUNT(fr.friend_1), u.first_name >> >> FROM users AS u >> >> LEFT

How are Dataframes partitioned by default when using spark?

2016-09-19 Thread Long, Xindian
How are Dataframes/Datasets/RDD partitioned by default when using spark? assuming the Dataframe/Datasets/RDD is the result of a query like that: select col1, col2, col3 from table3 where col3 > xxx I noticed that for HBase, a partitioner partitions the rowkeys based on region splits, can

Re: Full text query in Phoenix

2016-09-19 Thread Jean-Marc Spaggiari
HBase + Lily Indexer + SOLR will do that very well. As James said, Phoenix might not help with the full time. Google for that and you will find many pointers for web articules or even books. JMS 2016-09-19 9:05 GMT-04:00 Cheyenne Forbes : > Hi James, > > Thanks

Re: Using COUNT() with columns that don't use COUNT() when the table is join fails

2016-09-19 Thread Michael McAllister
This is really an ANSI SQL question. If you use an aggregate function, then you need to specify what columns to group by. Any columns not being referenced in the aggregate function(s) need to be in the GROUP BY statement. Michael McAllister Staff Data Warehouse Engineer | Decision Systems

Re: Combining an RVC query and a filter on a datatype smaller than 8 bytes causes an Illegal Data Exception

2016-09-19 Thread Samarth Jain
Kumar, Can you try with the 4.8 release? On Mon, Sep 19, 2016 at 2:54 PM, Kumar Palaniappan < kpalaniap...@marinsoftware.com> wrote: > > Any one had faced this issue? > > https://issues.apache.org/jira/browse/PHOENIX-3297 > > And this one gives no rows > > SELECT * FROM TEST.RVC_TEST WHERE

Re: Combining an RVC query and a filter on a datatype smaller than 8 bytes causes an Illegal Data Exception

2016-09-19 Thread Kumar Palaniappan
The problem is when we have just 1 param in the rvc it works. but this one , for 2+ SELECT * FROM TEST.RVC_TEST WHERE (COLONE, COLTWO) IN ((1,2),(1,2)) AND COLTHREE=3; blows up. On Mon, Sep 19, 2016 at 3:58 PM, Kumar Palaniappan < kpalaniap...@marinsoftware.com> wrote: > No, I didnt. > > But

Re: Using COUNT() with columns that don't use COUNT() when the table is join fails

2016-09-19 Thread Maryann Xue
Thank you very much for your answer, Michael! Yes, what Cheyenne tried to use was simply not the right grammar. Thanks, Maryann On Mon, Sep 19, 2016 at 10:47 AM, Michael McAllister < mmcallis...@homeaway.com> wrote: > This is really an ANSI SQL question. If you use an aggregate function, >

Re: Combining an RVC query and a filter on a datatype smaller than 8 bytes causes an Illegal Data Exception

2016-09-19 Thread Kumar Palaniappan
No, I didnt. But wrapping up with the parenthesis, it worked. SELECT * FROM TEST.RVC_TEST WHERE (COLONE, COLTWO) IN ((1,2)) AND COLTHREE=3; SELECT * FROM TEST.RVC_TEST WHERE ((COLONE, COLTWO) IN ((1,2)) AND (COLFOUR=4)); On Mon, Sep 19, 2016 at 2:56 PM, Samarth Jain wrote:

Re: Phoenix + Spark + JDBC + Kerberos?

2016-09-19 Thread Jean-Marc Spaggiari
Thanks for the pointer to PHOENIX-3189 Josh. I don't think we are facing that. We will try to activate the debug mode on Kerberos and retry. Good idea! I will keep this thread updated if we find something... JMS 2016-09-15 17:39 GMT-04:00 Josh Elser : > Cool, thanks for