Regarding Excel Files and Ms Access File.

2016-04-04 Thread Sanjiv Kumar
Please tell me when or in which updated version of Apache Drill, you are going to give storage plugin for query Excel (.xsl) and Ms Access File (.accdb).

Regarding Join Query Problem

2016-04-04 Thread Sanjiv Kumar
I want to join two table using same storage plugin. But One Of the Column showing null value. I am using this query:- SELECT T2.ID AS T_ID,T1.ID AS T1_ID ,T1.ProviderID AS ProviderID, T1.Name AS Name, T2.Code AS Code FROM SqlServer.StuentDetails.P21.Class1 AS T1 right outer JOIN SqlServer.Stuen

Re: Regarding Excel Files and Ms Access File.

2016-04-04 Thread Magnus Pierre
Hello Sanjiv, Can not answer the question you are asking for, but I recently opened a pull request towards Drill for a XML plugin which gives you ability to query Microsoft Excel Sheets saved as Excel 2004 XML Spreadsheets. You can find the code here if you want to try it: https://github.com/ma

Re: Regarding Join Query Problem

2016-04-04 Thread Magnus Pierre
Hello Sanjiv, It is perfectly valid to have two tables from the same plugin in a SQL statement in Drill. Not really sure you are doing something wrong. RIGHT or LEFT joins allow for returning null on the left / right side if the join condition is not met. If you only would like to have records

Query planner flow and optimizer rules

2016-04-04 Thread Lee Butts
Hi, I'm trying to add some custom optimizer rules for my storage format plugin and am having trouble understanding the process the query planner goes through when building logical and physical query plans. What I'd like to do is the following: 1) Add a rule which matches all filters and collects

Re: Query planner flow and optimizer rules

2016-04-04 Thread Jacques Nadeau
You should enable Calcite plan debugging. The instructions can be found by looking at the historical version of the Calcite howto before the logging framework changed in 1.5. [1] tracing section. Couple possibilities are you are not calculating distinct digests for your scan or your costs aren't r

Re: Query planner flow and optimizer rules

2016-04-04 Thread Jacques Nadeau
On your strategy, generally, rules should be independent and rel nodes should be immutable. I don't fully understand what you are trying to do but your description seems like a bit of an anti-pattern. You'll also need to figure out what phase or phases should include your rules. On Apr 4, 2016 9:5