Use this glossary when stages, pipes, streams, tasks, dynamic tables, sharing, and warehouse terms start to blur together. Keep it beside the cheat sheet and resources, not in place of scenario practice.
High-yield terms
| Term |
Short meaning |
Why it matters on DEA-C02 |
| Stage |
named location Snowflake can use for loading or unloading data |
core ingest boundary term |
| File format |
Snowflake object that defines how staged files are parsed |
core loading term |
COPY INTO |
explicit load command for bringing staged data into tables |
core batch-ingest term |
| Snowpipe |
automated staged-file ingest into Snowflake |
core unattended-ingest term |
| Snowpipe Streaming |
lower-latency streaming ingest path into Snowflake |
key near real-time ingest term |
| Dynamic table |
managed refreshed table defined by a query and target freshness |
key managed-transformation term |
| Stream |
Snowflake change-data-capture object |
key CDC term |
| Task |
Snowflake scheduled or triggered unit of work |
key orchestration term |
| Secure Data Sharing |
governed live data access for another Snowflake account |
key delivery term |
| Replication |
copying account or object state for continuity or cross-region needs |
key continuity term |
| Failover |
continuity action path when primary environment is unavailable |
key recovery term |
| Warehouse |
Snowflake compute resource for queries and data-engineering workloads |
key compute-fit term |
| Query profile |
Snowflake execution analysis view for diagnosing bottlenecks |
key performance-evidence term |
| Row access policy |
policy controlling which rows a consumer can see |
key governed-delivery term |
| Masking policy |
policy controlling how sensitive values are hidden or transformed |
key data-protection term |
Commonly confused pairs
| Pair |
Keep this distinction clear |
| stream vs task |
change capture versus execution scheduling |
| dynamic table vs stream-plus-task |
managed refresh versus explicit CDC plus orchestration |
| Snowpipe vs Snowpipe Streaming |
automated staged-file ingest versus lower-latency streaming write path |
| sharing vs replication |
live governed delivery versus copied continuity pattern |
| warehouse size vs performance diagnosis |
compute choice versus evidence-based tuning |
| row access policy vs masking policy |
row visibility versus value protection |
If three terms blur together
| Cluster |
Fast separation |
stage / COPY INTO / Snowpipe |
where data lives before load, how you load it explicitly, and how you automate staged-file ingest |
| stream / task / dynamic table |
capture changes, run work, or let Snowflake manage refresh |
| share / replication / failover |
deliver live data, copy state, or recover continuity |
| warehouse / query profile / activity history |
run compute, inspect bottlenecks, or inspect workload evidence |
One-sentence memory hooks
- If the question is about change capture, think stream before task.
- If the question is about scheduled execution, think task before stream.
- If the question is about managed refresh, think dynamic table before building CDC by hand.
- If the question is about live governed delivery, think sharing before copying.