I am trying to fetch a count of total columns for a list of individual tables/views from Impala from the same schema.
however i wanted to scan through all the tables from that schema to capture the columns in a single query ?
i have already performed a similar excercise from Oracle Exadata ,however since i a new to Impala is there a way to capture ?
select owner, table_name as view_name, count(*) as counts
from dba_tab_cols /*DBA_TABLES_COLUMNS*/
where (owner, table_name) in
(
select owner, view_name
from dba_views /*DBA_VIEWS*/
where 1=1
and owner='DESIRED_SCHEMA_NAME'
)
group by owner ,table_name
order by counts desc;
In Hive v.3.0 and up, you have INFORMATION_SCHEMA
db that can be queried from Hue to get column info that you need.
Impala is still behind, with JIRAs IMPALA-554 Implement INFORMATION_SCHEMA in Impala and IMPALA-1761 still unresolved.