Preparing Practice Exam…
Please wait a moment while the system generates the exam questions, answers, and study session.
Please wait a moment while the system generates the exam questions, answers, and study session.
SnowPro Core Certification · Snowflake
Select one answer.
Written by other learners. Comments can disagree with the marked answer — check the official documentation when in doubt.
Answer: Apache Iceberg tables (in Snowflake, usually Snowflake-managed Iceberg tables). Here’s how they meet each requirement: ACID transactions: Iceberg’s table format uses snapshot-based metadata, so every write is atomic and readers always see a consistent snapshot. Schema evolution without rewriting data: Iceberg tracks columns by ID rather than by name or position. You can add, drop, rename, or reorder columns, or widen types, without rewriting existing data files. Efficient querying on large datasets: Iceberg stores data in open Parquet files with rich metadata (manifests and partition and column statistics). This allows partition and file pruning, and Snowflake’s engine can query the tables with near-native performance. The “lakehouse” keyword is the main clue. Iceberg tables keep data in open format in your own cloud storage (through an external volume) and still behave like full Snowflake tables. Why the usual distractors are wrong External tables are read-only, so they offer no ACID writes, and they perform worse because they query files directly with limited metadata. Standard (native) Snowflake tables do support ACID, schema changes, and fast queries. However, the data sits in Snowflake’s proprietary format, which doesn’t fit a lakehouse with open storage. Hybrid tables are built for Unistore/OLTP workloads, meaning low-latency single-row operations, not lakehouse analytics. Transient and temporary tables are about persistence and Fail-safe behavior, which isn’t what this question tests. Related exam points Snowflake-managed Iceberg tables use Snowflake as the catalog and support full read/write. Externally managed Iceberg tables use a catalog such as AWS Glue or Snowflake Open Catalog (Polaris), and Snowflake’s access to them is typically read-only. Both kinds need an external volume to point to the cloud storage location.