Reputation: 525
Say there is a table A that is a time-partitioned table. To reduce cost of queries, my organization created a view of B, which queries only last 3 days of data. This view B is described in legacy SQL. Next month, new applications will be connecting to BigQuery to read the B data via an ODBC connection? Below are couple of options of which, which could be the best option?
A - Creating a new view C over view B using standard SQL
B - Creating a new view over table A using standard SQL
Upvotes: 0
Views: 630
Reputation: 59175
B is the better option:
This because you can't call a #legacySQL view from a query using #standardSQL. So the best road will be to create a similar view, but this time with #standardSQL.
Upvotes: 2