Optimizing Skewed Joins | Dataford Interview Questions

Given two large tables with highly skewed distribution keys, how would you optimize a complex SQL JOIN operation to avoid performance bottlenecks for Roku-scale datasets?

Using the supplied PostgreSQL tables, write the query that returns the requested result while demonstrating a skew-aware execution strategy. Assume content IDs 1 and 2 are identified as heavily skewed keys.

Output

One row per active content item with qualifying playback events and a matching channel.
Columns: content_id, title, channel_name, total_watch_seconds, and event_count.
Include playback events from January 2025 only, exclude null content IDs, and sort by total watch seconds descending, then content ID ascending.

Schema

viewing_events

Column	Type	Description
event_idPK	BIGINT	Unique playback event identifier
content_id	INT	Content item associated with the event
viewer_id	BIGINT	Viewer identifier
event_date	DATE	Date on which the event occurred
event_type	VARCHAR(30)	Playback event classification
watch_seconds	INT	Number of seconds watched

content_catalog

Column	Type	Description
content_idPK	INT	Unique content identifier
title	VARCHAR(150)	Content title
channel_id	INT	Owning channel identifier
is_active	BOOLEAN	Whether the content is currently active

channels

Column	Type	Description
channel_idPK	INT	Unique channel identifier
channel_name	VARCHAR(100)	Display name of the channel

Your question is Optimizing Skewed Joins. Start with the requirements and the three tables in the Question tab.

Run and submit as often as you like. When you're ready, talk me through your approach or go straight to the code.

Problem