Re: Syntax error in Kylin - even though query works in Hive

[Yiming Liu]

There is a slide introducing how Kylin works.
http://www.slideshare.net/XuJiang2/kylin-hadoop-olap-engine

2016-09-08 12:20 GMT+08:00 Something Something :

> It worked. Thanks. It's pretty fast as well.
>
> So now if you can explain how Kylin runs this in the background that would
> be great -:) What I mean is - in this query I am grouping by:
>
> group by part_dt, kylin_sales.LEAF_CATEG_ID, LSTG_FORMAT_NAME,
> LSTG_SITE_ID, META_CATEG_NAME
>
> Has Kylin already pre-built a "5 dimensional" array with all aggregates
> such as sum(price), max(price) etc?
>
> OR -
>
> Does it build HBase table as follows?
>
> Key:  (Dim1, Dim2. DimN).
> Value: Array of all aggregations (SUM, MAX etc)
>
> & when a query is triggered combines the Dimensions that are not
> specified?  Or some other way?
>
> Just want to understand how it behaves internally. Thanks.
>
>
>
>
> On Wed, Sep 7, 2016 at 8:11 AM, ShaoFeng Shi 
> wrote:
>
>> DEFAULT is a reserved keyword in Apache Calcite (which acts as Kylin's
>> SQL parser) ; when use it in SQL, please add double quote ("DEFAULT") to
>> escape;
>>
>> 2016-09-07 13:17 GMT+08:00 Something Something 
>> :
>>
>>> Here's a query that works in Hive, but gives a syntax error in Kylin.
>>> Any idea why?
>>>
>>> select part_dt, kylin_sales.LEAF_CATEG_ID, LSTG_FORMAT_NAME,
>>> LSTG_SITE_ID, META_CATEG_NAME, sum(price), max(price), min(price) from
>>> kylin_sales INNER JOIN DEFAULT.KYLIN_CATEGORY_GROUPINGS as
>>> KYLIN_CATEGORY_GROUPINGS ON KYLIN_SALES.LEAF_CATEG_ID =
>>> KYLIN_CATEGORY_GROUPINGS.LEAF_CATEG_ID AND KYLIN_SALES.LSTG_SITE_ID =
>>> KYLIN_CATEGORY_GROUPINGS.SITE_ID group by part_dt,
>>> kylin_sales.LEAF_CATEG_ID, LSTG_FORMAT_NAME, LSTG_SITE_ID, META_CATEG_NAME
>>> order by part_dt, kylin_sales.LEAF_CATEG_ID, LSTG_FORMAT_NAME,
>>> LSTG_SITE_ID, META_CATEG_NAME;
>>>
>>>
>>> Error message:
>>>
>>> Encountered "DEFAULT" at line 1, column 156. Was expecting one of:
>>>  ...  ...  ...
>>>  ...  ... "LATERAL"
>>> ... "(" ... "UNNEST" ... "TABLE" ...
>>>
>>
>>
>>
>> --
>> Best regards,
>>
>> Shaofeng Shi
>>
>>
>


-- 
With Warm regards

Yiming Liu (刘一鸣)

Re: Use Apache HUE with Kylin

[Alberto Ramón]

>From Hue MailList Link

(Romain Rigaux) :

Great guide!

I updated to refer to it:
http://gethue.com/sql-editor/
http://gethue.com/custom-sql-query-editors/

Thanks!

PS: hope the metadata bug gets picked up at some point ;)


BR, Alberto Ramón

2016-08-30 11:09 GMT+02:00 Luke Han :

> Thanks Alberto, your contribution will benefit many people:)
>
>
>
>
> Best Regards!
> -
>
> Luke Han
>
> On Mon, Aug 29, 2016 at 11:38 PM, Alberto Ramón  > wrote:
>
>> @Luke Han:  fell free to add where you want
>> (I think the integration with HUE, will be very interested due to
>> Notebooks, The people use it a lot in reports of BI or ML, with MarkDown )
>>
>> About Kylin + Hue:
>>   - Hue 3.10 & 3.11, have the same issues. But in the next version
>> ¿3.12? Will solve some bug, and the integration will be better and easier
>>   (Hue 3.10 is in Cloudera 5.8, and 3.11 have less than 10 days old ¡¡,
>> --> It's a very new feature)
>>   - Hue Hue 3228 , can be
>> VERY interested but is only in the road map
>>
>> About Kylin + Other programs:
>>   - I'm testing with other programs, when I have final results, I will
>> put on GitHub and MailList
>>
>> BR, Alberto
>>
>> 2016-08-29 14:46 GMT+02:00 Luke Han :
>>
>>> Hi Alberto,
>>> Would you mind to share more to create one how to wiki and submit
>>> patch to doc branch? I think this will help a lot for users who want
>>> leverage Hue with Kylin.
>>>
>>> Thanks.
>>> Luke
>>>
>>>
>>> Best Regards!
>>> -
>>>
>>> Luke Han
>>>
>>> On Sun, Aug 28, 2016 at 12:11 PM, hongbin ma 
>>> wrote:
>>>
 ​thanks Alberto!​




 --
 Regards,

 *Bin Mahone | 马洪宾*

>>>
>>>
>>
>