Study SnowPro COF-C02 Stages and Snowpipe: key concepts, common traps, and exam decision cues.
On this page
This lesson matters because many data-movement questions reduce to four ideas: where the files are, how they are parsed, what command performs the load, and whether the ingest pattern is batch or continuous.
Load-path chooser
Requirement
Strongest first fit
files already staged for batch ingestion
COPY INTO
continuous file arrival with low management overhead
Snowpipe
parsing CSV, JSON, or Parquet correctly
file format object
knowing where files are stored
stage
Read load questions in the right order
Ask this first
Why it matters
where are the files?
that points to the stage boundary
how should Snowflake interpret the files?
that points to the file-format object
is the load one-off or continuous?
that separates COPY INTO from Snowpipe logic
Snowflake often gives distractors from all three layers. Strong answers classify the requirement first.
Decision order that usually wins
Ask where the files are.
Ask how Snowflake should parse them.
Ask whether the ingest is batch or continuous.
Use COPY INTO for explicit staged batch loads.
Use Snowpipe only when the arrival pattern truly demands continuous ingest.
Common traps
Trap
Better rule
treating the stage like the parsing rule
stage and file format are different objects
using Snowpipe to answer every ingest question
Snowpipe is strongest when the arrival pattern is continuous
debugging SQL transforms before the files load correctly
confirm stage, format, and load action first
Scenario triage
Scenario clue
Stronger answer shape
“files already landed and need batch ingest now”
COPY INTO
“new files keep arriving and low-touch ingest is preferred”
Snowpipe
“CSV loads are misread or column parsing is wrong”