What causes Snowpipe Ingestion Lag / Backlog?

The incoming file volume exceeds the throughput Snowpipe can process with current warehouse resources. Files are being staged faster than Snowpipe's polling interval can detect and process them. The Snowpipe pipe was paused (manually or due to an error) and files accumulated while paused. Error notifications are not configured, so failed files silently accumulate in the backlog. Large individual files (>250 MB) slow Snowpipe processing; Snowflake recommends files of 100–250 MB

How do I fix Snowpipe Ingestion Lag / Backlog?

Check the pipe status: `SELECT SYSTEM$PIPE_STATUS(' ')` — look at pendingFileCount and windowStart.. If the pipe is paused, resume it: `ALTER PIPE RESUME`.. Configure Snowpipe error notifications (SNS, Pub/Sub, or Event Grid) to receive alerts when files fail ingestion.. Split large files into 100–250 MB chunks before staging to maximize Snowpipe throughput.. For sustained high volume, consider switching to Snowpipe Streaming (SDK-based) which provides higher throughput and lower latency.. Review the `COPY_HISTORY` view to identify files that failed silently: `SELECT * FROM TABLE(INFORMATION_SCHEMA.COPY_HISTORY(TABLE_NAME=>' ', START_TIME=>DATEADD('hour',-24,CURRENT_TIMESTAMP()))) WHERE STATUS != 'Loaded'`.

Medium severitydata source

Power BI Refresh Error:
Snowpipe Ingestion Lag / Backlog

What does this error mean?

Snowpipe is not processing files as fast as they arrive, causing a growing backlog. Files are queued but not ingested within the expected SLA, resulting in stale data in downstream tables.

Common causes

1The incoming file volume exceeds the throughput Snowpipe can process with current warehouse resources
2Files are being staged faster than Snowpipe's polling interval can detect and process them
3The Snowpipe pipe was paused (manually or due to an error) and files accumulated while paused
4Error notifications are not configured, so failed files silently accumulate in the backlog
5Large individual files (>250 MB) slow Snowpipe processing; Snowflake recommends files of 100–250 MB

How to fix it

1Check the pipe status: `SELECT SYSTEM$PIPE_STATUS('<pipe_name>')` — look at pendingFileCount and windowStart.
2If the pipe is paused, resume it: `ALTER PIPE <pipe_name> RESUME`.
3Configure Snowpipe error notifications (SNS, Pub/Sub, or Event Grid) to receive alerts when files fail ingestion.
4Split large files into 100–250 MB chunks before staging to maximize Snowpipe throughput.
5For sustained high volume, consider switching to Snowpipe Streaming (SDK-based) which provides higher throughput and lower latency.
6Review the `COPY_HISTORY` view to identify files that failed silently: `SELECT * FROM TABLE(INFORMATION_SCHEMA.COPY_HISTORY(TABLE_NAME=>'<table>', START_TIME=>DATEADD('hour',-24,CURRENT_TIMESTAMP()))) WHERE STATUS != 'Loaded'`.

Frequently asked questions

How do I know if files failed Snowpipe ingestion silently?

Query INFORMATION_SCHEMA.COPY_HISTORY for STATUS != 'Loaded'. Also set up Snowpipe error notifications — without them, failed files drop silently and data gaps are invisible until downstream consumers notice.