Export Oracle to GCS in Parquet
Fast, parallel data export with zero intermediate storage
Terminal
.\FastBCP.exe `
--connectiontype "oraodp" `
--server "host.domain | host.domain,port | host.domain,port/service" `
--database "tpch" `
--trusted `
--sourceschema "tpch10" `
--sourcetable "orders" `
--query "SELECT * FROM tpch10.orders" `
--directory "gs://rootbucket/fastbcpexport/raw/{sourcedatabase}/{sourceschema}" `
--fileoutput "{sourcetable}.parquet" `
--method "Ntile" `
--distributekeycolumn "o_orderkey" `
--paralleldegree -2 `
--merge false `
--runid "runidfromcaller"Source - Oracle Database
Oracle Database is the world's most widely used enterprise database. FastBCP leverages the ODP.NET driver for maximum performance during Oracle exports.
Features:
- •Native Oracle ODP.NET driver
- •Full support for Oracle-specific types
- •Optimized for large volumes
Parallel Method - Ntile
Divides data into N equal partitions based on a numeric column.
Requirement: Requires a numeric distribution column
Available parallel methods with Oracle:
Output Format - Apache Parquet
Parquet is a columnar file format optimized for analytical processing. FastBCP exports to Parquet with compression and type preservation, ideal for data lake architectures.
Features:
- •Columnar storage for efficient analytics
- •Integrated compression (Snappy, Gzip)
- •Full data type preservation
- •Optimized for big data tools (Spark, Hadoop)
Destination - Google Cloud Storage
Google Cloud Storage provides unified object storage. FastBCP uploads files to GCS buckets with parallel chunked uploads for fast transfers.
Storage Type:
Cloud Object Storage
Features:
- •Parallel chunked upload
- •Service account authentication
- •Bucket and prefix configuration
- •Multi-regional storage classes
