A shared city AI platform in Cosmos DB must partition by city, department, and document type together. What feature helps?
Select an answer to reveal the explanation.
Short Explanation
Hierarchical partition keys are nested city → department → docType folders—Cosmos can fan traffic across the tree instead of jamming every office into one overflowing drawer.
Full Explanation
A shared city AI platform in Cosmos DB that must partition by city, department, and document type together benefits from hierarchical partition keys spanning those dimensions. Hierarchical keys distribute multi-department municipal AI data along the natural hierarchy instead of collapsing everything into one hot logical partition that throttles under council-week load. Ignoring partitioning and placing all departments into one logical partition forever concentrates RU and storage risk and fails when parks, permits, and clerk traffic spike together. Partitioning only by random UUID with no city or department hierarchy spreads data but loses hierarchical query and fan-out patterns the platform needs for scoped municipal retrieval. Requiring a separate Cosmos account for every single PDF uploaded by staff creates unmanageable account sprawl and does not solve in-account partitioning for shared AI services. Exam caveat: hierarchical partition keys help when the access pattern aligns with the hierarchy; poorly chosen levels still create hot partitions that operators must detect. Operationally, define city → department → docType keys to match query filters, review partition RU heatmaps by department, and adjust key design if one office dominates traffic.