Your team runs a complex analytical query daily that processes terabytes of data. Recently, after running for
20 minutes, the query fails with a "Resources exceeded" error. You need to resolve this issue. What should you do?
Correct Answer: D
In BigQuery, the "Resources exceeded" error (specifically during execution) typically indicates that the query's resource demands (CPU/Memory/Shuffle) have surpassed the limits of the shared on-demand slot pool.
* Slot Reservations: On-demand pricing uses a shared pool of slots with a soft cap (typically 2,000 slots). For massive, complex queries processing terabytes and running for long durations, this pool may be insufficient or subject to "load shedding" during peak times. Moving to Capacity-based pricing (Slots) allows you to reserve a dedicated number of slots (e.g., 500, 2,000, or more) that are exclusively yours, providing the sustained compute power needed for heavy analytical jobs.
* Correcting other options:
* A: API request quotas (e.g., queries per second) are unrelated to the internal compute resources required to execute a single massive query.
* B: While query optimization (removing ORDER BY, etc.) can help, if the logic is correct but simply too large for the shared pool, syntax analysis won't fix the underlying resource constraint.
* C: There is no "maximum table size limit" that causes an execution-time resource error; BigQuery supports petabyte-scale tables.
Reference: Google Cloud Documentation on BigQuery Troubleshooting:
"Resources exceeded during query execution: This error occurs when a query uses too many resources (CPU, memory, or shuffle). This is common for queries that are very complex or join large datasets... To resolve this, you can: 1. Optimize your query... 2. Switch to capacity-based pricing (reservations) to ensure your jobs have a dedicated number of slots and are not impacted by the shared on-demand pool's limits." (Source: Troubleshoot query issues)
"On-demand pricing is subject to a default slot quota... For workloads that require more predictability or higher scale, slot commitments and reservations provide dedicated capacity." (Source: BigQuery pricing models)