I have some metadata tables and fact tables holding ad impression data for 100 clients. My clients want ad-hoc reporting on top of canned reports on their own data. I want to limit #of tables(to save storage cost) and thus do not want to create separate tables for each one of my client (holding only their own data). Creating a single table is advantageous but that raises lot of security concerns. Is there any way i can protect data-access so that clients cannot see each other's data ? As an advertiser agency (managing ad campaigns for my advertisers/clients), i want to run cross-client trend/pattern analysis and that is the only reason why i have decided to create only one table holding data for all of my 100 advertisers. Single table also removes query-join issues. I am creating Daily partitions(10TB each).I do want to "limit" data processing volume (to save query cost) when a particular client scans only his own data from partition holding data from other 99 clients. Could i seek help on my data model as well as how to ensure best security for data access in BigQuery ?
0
votes
1 Answers
0
votes
Why would you want to limit the number of tables? As far as I can tell, the BigQuery storage pricing only looks at the number of bytes stored.
For doing cross-client analysis, you can use wildcards to query up to 1000 tables.
