26
Accepted answer
Document the grain decision, most BI bugs turn out to be grain bugs.
Asked to estimate storage for a logging system and got stuck trying to recall exact figures for average log line size and compression ratios. Interviewer seemed to want a specific number, not just an approach. Is memorizing benchmark numbers actually expected?
Accepted answer
Document the grain decision, most BI bugs turn out to be grain bugs.
Small nit: the broadcast hint gets ignored once the table is over threshold, check the UI to confirm.
Not exact memorized numbers. What's expected is reasonable, stated assumptions, like 'assume 200 bytes per log line uncompressed, roughly 5 to 10x compression with a columnar format,' and then doing the arithmetic cleanly. The assumption itself matters less than showing you know which assumptions the calculation depends on.
Prefer a staging table plus validation gate before promoting to prod tables.
Tried that, partial improvement but lag still spikes on redeploy.
How do you handle backfill without duplicating rows?
Worth measuring the serialized size before choosing broadcast.
Sign in to reply.
© 2026 Lakebench, operated by Hunnurji Rao. Bengaluru, Karnataka, India.
No cluster. No install. Just the tab.