How to swap a table underlying a pinboard with a newly imported table ?
I have a request from my user community. They want to be able to rapidly spin up the TS development process in Thoughtspot by creating a table "A" (any way they can), (import from csv or TQL) ; then create a view on top of the table (wksht), then a pinboard on top of that.
The catch is that they then want to be able to create a (new and improved version of the) table (Table "B") in an Oracle data-warehouse using the IT deployment process ( which admittedly is a slower process ). Then using the "Add data source" Import the newly created table B and swap out the underlying table A originally used for the pinboard development with table B.
Is anything like this possible?
Does anyone else have this need?
Steven Cover it's not possible to easily swap the tables out but you could try using Data Connect to load data from the new Table B into existing Table A. See https://help.thoughtspot.com/Integration_Guides/Data_Connect_Guide_4.2/010/030/010 (4.4) https://docs.thoughtspot.com/4.4/data-connect/setup/about-changing-DDL.html although the user would need access to TQL. The column names and order for the new table will need to exactly match the existing table. You can also find user uploaded tables using TQL - they are stored in the "FalconUserDataDataBase" database ("FalconUserDataSchema" is the schema). User uploaded tables are given generated names and the easiest way I know to find that generated name is by downloading the Modeling file and checking the PhyTableName column, which would require Admin access.
I do have Admin access and access to the TQL. I am familiar with SQL and have "tried" some tql commands and they seem similar so that's all good.
So if I understand I can create a table in the db using TQL. Load some sample data into the table I just created via Import from CSV or EXCEL. Do some development work.
Then do the work in my data warehouse setting up the final version of said table. Truncate the dev/test sample data in the thoughtspot table I created using TQL.
Then finally use Data Connect to load up the final groomed data from my data warehouse into the table.
Is this what you are saying?
Hi Vanessa McAfee ,
I uploaded a csv and created pinboards on it
Now want to re-upload csv with additional columns, in place of existing csv in thoughtspot without editing pinboards.I tried locating "PhyTableName" ID of my table , from modeling file, but don't know where to exactly write a query to alter table.
cant use physical table name in the query and the query with actual file name is showing error.
I would like to add that we have now had multiple requests from our customers for this ability as well. In our case we would not be appending data to the data set but the data set values change over time. AKA - the data on a set of suppliers - number of suppliers stays the same but the details about them may change. Our users do not have access to Data Connect.