-
shopify.engineering/reducing-bigquery-costs The superficial takeaway is 1. use clustering 2. Avoid the SELECT* statement 3. Partition your tables 4. Don’t run queries to explore or preview data The real takeaway, I think, is 1. experiment in the small 2. think carefully about requirements at (1/2)
Mood +1 🙂
-
scale 3. understand the business constraints (e.g. cost) 4. know your tools well enough to take into account what you learned from the experiment, the requirements, and constraints 5. experiment some more to prove your thesis (2/2)Permalink
