Remaining 5 file types are need to be tested for "Data Catalog".Ģ.
Test Cases are added to the test rail: These are the Scenarios which are pending due to S3 filesġ. Create External Table, Alter External Table, Drop External Table, Select top 1000 rows, External Table Properties, Script object (to window as, to new window as, to file as)Ĥ. Create External Schema (data catalog / Hive Metastore), Drop External Schema (with & w/o delete database), External Schema Properties, Script object (to window as, to new window as, to file as)ģ.
Redshift external table manual#
I have covered all the possible scenarios regarding manual testing those are as follows:-Ģ. While alter the table when we add any column, "Not Null" value getting generated in preview SQL tab, which is not getting generated while creating table. (refreshing External Databases node taking around 5-6 minutes) screenshot6īhushan will attach the screenshots for these steps. The external database 'bhushan2' is also dropped. Click on menu 'Drop External Schema' for 'bhushan2schema1', in the dialog I checked the option 'Drop External Database' and click on Drop button. Now database 'bhushan2' have two external schemas 'bhushan2schema' and 'bhushan2schema1'. The external schema 'bhushan2schema1' created. Created another external schema 'bhushan2schema1' under database 'bhushan2'. (refreshing External Databases node taking around 5-6 minutes). The external database 'bhushan2' and schema 'bhushan2schema' created successfully. Created external schema 'bhushan2schema' with option CREATE EXTERNAL DATABASE IF NOT EXISTS database name 'bhushan2'. It is working fine, I tested with following steps:ġ. This is important because we can't directly drop the database. If unchecked, just change the cascade to restrict. Have it checked by default and add the suffix cascade in the ddl. We should also have a checkbox in the gui external schema drop for cascade. My guess is that it senses the schema and doesn't think the database is empty.ĭROP SCHEMA "test" DROP EXTERNAL DATABASE CASCADE For the gui drop, you use the suffix restrict(default) which doesn't seem to drop the database. If I use the drop with cascade in the QA window, it also drops the database. If I drop an external schema using the gui schema tree drop with the drop database checked, it doesn't seem to drop the database in the schema tree. Hmm!! Have to think about this?Īn observation. This looks problematic and does not work.If you happen to select the internal database "tom" in the QA window and then execute SELECT * FROM tom.student_data which is an external table, it errors with a not found. I don't know how you make the distinction between and a "tom" internal database and a "tom" external database. When doing a script table to new window, we get this error message. We probably should just use the default internal database and schema for the new window instead of trying to switch to the external one. Can we either eliminate it or gray it out for external tables whichever is easier. I don't see that we can have those in the ddl. Table Properties now has an extra line Would you please have a look at it?Īnother thing I noticed was that for Alter/Create, the null column is available. Please test it and clean it up - empty catch.
Redshift external table how to#
I didn't look at all the new code that you checked in but I did add a quick fix(hack) to demonstrate how to work around this. It kind of tricks it and I didn't see any problems when doing this. Notice in the node classes like RedshiftExtSchemasNode, I use ConnectionProperties connSettings = getConnectionProperties() to get the original connection and database information. You cannot use our connection scheme to connect to them. When making a connection, remember that external databases are place holders. There are no internal databases named harish, harish1.etc. The reason databases tom and tomtest work correctly with the new external schema gui is that there are already internal database's with that name.
0 kommentar(er)
