Network instability causing repeated failures during multipart result download from S3/Azure Blob/GCS. Proxy or load balancer timing out long-lived download connections. Insufficient bandwidth or high packet loss on the network path to cloud storage. The presigned URL for the result chunk expired before download completed. Client-side memory pressure causing slow reading and subsequent timeouts

Reduce the result set size: add LIMIT, WHERE clauses, or paginate the query to fetch fewer rows.. Check network stability between the client and the cloud storage region (S3/Azure/GCS).. Increase the JDBC network timeout: add networkTimeout= to the JDBC connection URL.. If behind a proxy, ensure the proxy supports long-lived connections for large data downloads.. Consider using Snowflake's unload (COPY INTO @stage) for very large result sets and download separately.

Medium severityexecutionSnowflake →

Power BI Refresh Error:
JDBC 6

What does this error mean?

The JDBC driver exhausted all retry attempts while downloading a large query result chunk from Snowflake's cloud storage. The connection was established and the query executed successfully, but result retrieval failed.

Common causes

1Network instability causing repeated failures during multipart result download from S3/Azure Blob/GCS
2Proxy or load balancer timing out long-lived download connections
3Insufficient bandwidth or high packet loss on the network path to cloud storage
4The presigned URL for the result chunk expired before download completed
5Client-side memory pressure causing slow reading and subsequent timeouts

How to fix it

1Reduce the result set size: add LIMIT, WHERE clauses, or paginate the query to fetch fewer rows.
2Check network stability between the client and the cloud storage region (S3/Azure/GCS).
3Increase the JDBC network timeout: add networkTimeout=<ms> to the JDBC connection URL.
4If behind a proxy, ensure the proxy supports long-lived connections for large data downloads.
5Consider using Snowflake's unload (COPY INTO @stage) for very large result sets and download separately.

Frequently asked questions

Is there a result size limit for JDBC queries?

There is no hard limit, but very large result sets (millions of rows) are slow and fragile. Use COPY INTO for bulk exports — it is more reliable and faster than JDBC for large data transfers.

What is a result chunk in Snowflake?

Large query results are split into compressed chunks stored in cloud storage (S3/Azure/GCS). The JDBC driver downloads these chunks in parallel. Error JDBC 6 means too many chunk download attempts failed.

Official documentation: https://docs.snowflake.com/en/user-guide/client-connectivity-troubleshooting/error-messages