I have Tableau connected to Cloudera Impala data. The table I am reading from Metastore keeps on updating so, when I want the viz to update (pressing F5) I have to go to Impala in Cloudera and execute "INVALIDATE METADATA" before refreshing the viz. I know this can be done using the connection to Hive Server 2, but it takes a long time to execute the query. My question is there anyway to automatically execute "validate metadata" and "Refresh" queries in Impala Cloudera?
Thanks,
Amr
Unfortunately Impala does not offer a mechanism to automatically update metadata yet. We are working on this for the future, though we haven't committed such functionality to a particular release yet.
In the meantime, there may be some things you can do to make this easier. Firstly, how is the table being updated? Are there new data files? If so, you can run REFRESH which is faster. Also, be sure to only INVALIDATE/REFRESH the specific table, e.g. REFRESH my_table. See the documentation for [INVALIDATE METADATA][1] and REFRESH for more information.
Is there some ETL process in the background? Some users modify their workflow to issue a REFRESH command to impala after updating a table, e.g.
in a script adding impala-shell -i <impalad-host> -q "REFRESH my_table"
