mongo2pg¶
mongo2pg is a command-line tool that inspects MongoDB collections, infers
their structure, generates PostgreSQL DDL, exports relationally-shaped CSV
data, and produces HTML reports to support MongoDB-to-PostgreSQL migrations.
Why mongo2pg?¶
Moving from MongoDB to PostgreSQL is usually not blocked by raw data volume. It is blocked by shape:
- embedded documents become dependent tables
- arrays become one-to-many expansions
- mixed field types become type-mapping decisions
- very long nested names become PostgreSQL identifier problems
mongo2pg helps you answer those questions before you load data into
PostgreSQL.
Key Features¶
| Feature | Description |
|---|---|
| Schema inference | Samples MongoDB collections and writes JSON schema-like structure plus stats |
| Project workflow | init creates a repeatable migration project with config, source, schema, data, and report folders |
| PostgreSQL DDL generation | infer refreshes PostgreSQL CREATE TABLE statements under schema/tables/ when used with a project config |
| Per-collection PostgreSQL schemas | Each collection is deployed into its own PostgreSQL schema by default |
| CSV export | export expands MongoDB documents into .csv.gz files matching the generated SQL tables |
| PostgreSQL import | import creates PostgreSQL objects and loads exported .csv.gz files using TARGET_URI |
| HTML reporting | infer writes collection-level and schema-level reports, and report can regenerate them from existing files |
| Post-import validation | import writes reports/post_report.html automatically, and report --post-import can regenerate it later |
Typical Workflow¶
# 1. Create a migration project
mongo2pg init \
--project-base ./projects \
--project-name sample_airbnb \
--source-uri 'mongodb://user:pass@localhost:27017/?authSource=admin' \
--target-uri 'postgres://postgres:x@localhost:5432/postgres?sslmode=disable' \
--namespace sample_airbnb
# 2. Infer schemas, generate PostgreSQL DDL, and write the main reports
mongo2pg infer -c ./projects/sample_airbnb/config/sample_airbnb.toml
# 3. Export relational CSV files
mongo2pg export -c ./projects/sample_airbnb/config/sample_airbnb.toml
# 4. Create PostgreSQL objects and import the exported CSV files
mongo2pg import -c ./projects/sample_airbnb/config/sample_airbnb.toml
# 5. `import` already generated reports/post_report.html.
# Run this only if you want to regenerate it.
mongo2pg report -c ./projects/sample_airbnb/config/sample_airbnb.toml --post-import
Outputs¶
mongo2pg works from a project directory that separates concerns clearly:
<project>/
config/ project configuration (`SOURCE_URI`, `TARGET_URI`, `NAMESPACE`)
source/collections/ inferred collection schemas and stats
schema/tables/ generated PostgreSQL DDL
data/ exported `.csv.gz` files
reports/ HTML reports and schema diagrams
Screenshots¶
Cluster overview¶
Database drill-down¶
Schema diagram¶
Post-import validation¶
