BigQuery Alerts - Paradime

Radar’s BigQuery Data Alerts detect daily spend spikes and query anomalies across your BigQuery project. Enable and tune each alert on the BigQuery tab of the Radar Data Alerts page. See Configure Alerts for how to set thresholds and destinations.

Available alerts

Alert	Triggered when	Threshold	Example
Project Cost Anomaly	A project’s daily spend exceeds its 7-day average above the threshold.	% Change	`20`
User Cost Anomaly	A user’s daily spend exceeds their 7-day average above the threshold. Considers the top 10 users by spend.	% Change	`20`
dbt™ Cost Anomaly	A dbt™ model’s daily spend exceeds its 7-day average above the threshold. Considers the top 10 dbt™ models by spend.	% Change	`20`
Query Cost Anomaly	A specific query’s cost is more than its 7-day average above the threshold. Considers the top 10 queries by cost.	% Change	`20`
Small Query Waste Anomaly	Wasted bytes exceed a set percentage of total billed bytes in a day (waste = billed bytes minus processed bytes). Surfaces query inefficiencies, especially small queries with minimum billing.	% Change	`5`
High Shuffle Spill Anomaly	Disk-spilled shuffle bytes exceed a set percentage of total processed bytes for your top 10 largest queries. High spillage means performance problems and higher costs.	% Change	`20`
Query Runtime Anomaly	A query’s maximum runtime jumps significantly above its weekly average. Tracks your top 10 most expensive queries from yesterday. Longer runtimes may signal performance issues or data volume changes that raise costs.	% Change	`50`
Query Slot Usage Anomaly	A query’s maximum slot usage jumps significantly above its weekly average. Tracks your top 10 most expensive queries from yesterday. Higher slot usage signals performance issues, query complexity changes, or data volume increases that raise costs.	% Change	`50`

Each alert also has a Minimum value floor that must be met alongside the percentage change. See Configure Alerts.