What causes PIPE_EXECUTION_PAUSED?

Pipe was explicitly paused: ALTER PIPE my_pipe SET PIPE_EXECUTION_PAUSED = TRUE. Snowflake automatically paused the pipe after detecting a persistent error (e.g., file format mismatch, access denial). Account suspension due to credit limit caused all pipes to be paused. The pipe's IAM role or SAS token (for external stages) expired or was revoked. Pipe definition references a stage or file format that no longer exists

How do I fix PIPE_EXECUTION_PAUSED?

Check pipe status: SELECT SYSTEM$PIPE_STATUS('mydb.myschema.my_pipe').. Resume the pipe by running: `ALTER PIPE mydb.myschema.my_pipe RESUME`.. Inspect the error that caused auto-pause: the SYSTEM$PIPE_STATUS output includes an executionState and a reason field.. Verify the stage and file format referenced by the pipe still exist and the executing role has access.. After fixing the root cause, refresh any files that were not ingested during the pause: SELECT SYSTEM$PIPE_FORCE_RESUME('my_pipe').. Set up regular Snowpipe health checks — pauses do not raise automatic alerts unless you query SYSTEM$PIPE_STATUS.

High severityinfrastructureSnowflake →

Snowflake Error:
PIPE_EXECUTION_PAUSED

What does this error mean?

A Snowpipe is in a paused state and is not ingesting new files from the stage. This can happen due to explicit user action, a billing suspension, or Snowflake pausing a pipe after repeated errors.

Common causes

1Pipe was explicitly paused: ALTER PIPE my_pipe SET PIPE_EXECUTION_PAUSED = TRUE
2Snowflake automatically paused the pipe after detecting a persistent error (e.g., file format mismatch, access denial)
3Account suspension due to credit limit caused all pipes to be paused
4The pipe's IAM role or SAS token (for external stages) expired or was revoked
5Pipe definition references a stage or file format that no longer exists

How to fix it

1Check pipe status: SELECT SYSTEM$PIPE_STATUS('mydb.myschema.my_pipe').
2Resume the pipe by running: `ALTER PIPE mydb.myschema.my_pipe RESUME`.
3Inspect the error that caused auto-pause: the SYSTEM$PIPE_STATUS output includes an executionState and a reason field.
4Verify the stage and file format referenced by the pipe still exist and the executing role has access.
5After fixing the root cause, refresh any files that were not ingested during the pause: SELECT SYSTEM$PIPE_FORCE_RESUME('my_pipe').
6Set up regular Snowpipe health checks — pauses do not raise automatic alerts unless you query SYSTEM$PIPE_STATUS.

Frequently asked questions

Are files that arrived while the pipe was paused automatically ingested after resume?

New files arriving after resume are ingested automatically. Files that arrived during the pause may need to be manually triggered: use ALTER PIPE ... REFRESH or SYSTEM$PIPE_FORCE_RESUME to backfill missed files.

How do I monitor Snowpipe health in production?

Schedule a recurring query against SYSTEM$PIPE_STATUS for each production pipe. Alternatively, query SNOWFLAKE.ACCOUNT_USAGE.PIPE_USAGE_HISTORY to monitor file ingestion rates and detect drops.

Source · docs.snowflake.com/en/user-guide/data-load-snowpipe-intro