Odoo and other ERP workloads
Odoo only runs on PostgreSQL, so performance, backups and upgrades of the ERP depend directly on how the database is configured and maintained.
Technologies · PostgreSQL
PostgreSQL development for data that has to stay correct
Every Odoo system we deliver runs on PostgreSQL, and most of our custom applications do too. We design the schema, watch the slow queries, and make sure a backup can actually be restored.
What is PostgreSQL?
PostgreSQL development services cover designing, building, tuning and operating databases on PostgreSQL, the open-source relational database used by Odoo and many custom business applications. PostgreSQL enforces constraints and transactions strictly, supports JSONB documents, full-text search and geospatial data through PostGIS, and replicates to standby servers. The work ranges from modelling tables and indexes to setting up replication, point-in-time recovery and major-version upgrades.
Where PostgreSQL earns its place
Transactional business data
Orders, invoices, stock moves and payments where foreign keys, constraints and ACID transactions prevent silent inconsistencies.
Mixed relational and document data
JSONB columns hold flexible attributes or API payloads next to normal tables, with GIN indexes, so a separate document store is often unnecessary.
Reporting on live operational data
Window functions, CTEs, materialized views and read replicas let analysts query production data without slowing down users.
When we recommend something else
- A WordPress or WooCommerce site — WordPress is built for MySQL or MariaDB, and running it on anything else is unsupported territory.
- Very high-volume time-series or event analytics where a columnar store such as ClickHouse answers aggregate queries far more cheaply.
- A small internal tool with one user and no concurrency — SQLite is simpler to ship and back up.
- Caching, sessions or queues that need sub-millisecond reads — Redis or a message broker fits better than another database table.
PostgreSQL work we take on
Schema design for new applications
Normalized tables, constraints, indexes and migration scripts versioned alongside the application code.
Odoo database tuning
Configuration sized to the server, slow-query analysis with pg_stat_statements, index review on heavy custom modules, and vacuum settings for busy tables.
Replication and high availability
Streaming replicas for failover and read scaling, with automated failover through Patroni where the risk justifies it.
Backups, recovery and upgrades
pgBackRest with point-in-time recovery, scheduled restore tests, and major-version upgrades planned with pg_upgrade or logical replication.
A production PostgreSQL stack, layer by layer
L1
Application layer
Odoo workers or application services send queries through an ORM or hand-written SQL that we review for N+1 patterns.
L2
Connection pooling
PgBouncer keeps the number of real database connections small, with the pooling mode chosen to match the application's use of sessions.
L3
Primary server
Memory, WAL, checkpoint and autovacuum settings sized to the hardware and the write pattern, not left at defaults.
L4
Standby replicas
Streaming replication to one or more standbys for failover and for read-only reporting queries.
L5
Backup repository
Full and incremental backups plus continuous WAL archiving to separate storage, enabling recovery to a point in time.
L6
Monitoring
Slow queries, replication lag, bloat, disk growth and connection counts tracked with alerts before they become outages.
Ecosystem we work with
- PgBouncer
- pgBackRest
- Patroni
- PostGIS
- pgvector
- pg_stat_statements
- Logical replication
- pg_upgrade
- TimescaleDB
- Prometheus postgres_exporter
- Amazon RDS / Aurora
- Odoo ORM
What a well-run PostgreSQL gives you
Integrity enforced in the database
Foreign keys, unique and check constraints catch bad data even when an import script or integration bypasses the application.
Queries tuned from evidence
We read EXPLAIN ANALYZE plans and pg_stat_statements before adding indexes, so fixes target the queries that actually cost time.
Recoverable by design
Point-in-time recovery means an accidental bulk delete can be rolled back to the minute before it happened, if restores are tested.
No license lock-in
PostgreSQL is open source under a permissive license, so it runs the same on your own server, any cloud, or a managed service.
PostgreSQL or MySQL?
| Consideration | PostgreSQL | MySQL |
|---|---|---|
| Typical platform fit | Odoo, Django, custom ERP and data-heavy apps | WordPress, WooCommerce, Magento and many PHP apps |
| Data integrity | Strict by default; check constraints, exclusion constraints, transactional DDL | Strict with InnoDB and strict SQL mode; DDL is not transactional |
| Semi-structured data | JSONB with GIN indexes and rich operators | JSON type with generated-column and functional indexes |
| Extensions | PostGIS, pgvector, TimescaleDB, pg_partman and more | Fewer extensions; features come with server releases |
| Connection handling | Process per connection; a pooler such as PgBouncer is usually needed at scale | Thread per connection; handles many connections natively |
Related services, technologies and industries
Sources and further reading
Official documentation and guidance if you want to read more about this topic.
PostgreSQL questions
Why does Odoo use PostgreSQL?
Odoo is built only for PostgreSQL because its ORM relies on PostgreSQL features such as strict transactions, sequences and specific SQL behaviour. MySQL, SQL Server and other databases are not supported, so tuning PostgreSQL is part of tuning Odoo.
Can you speed up a slow Odoo database?
Often, yes, but the cause has to be found first. Slowness usually comes from untuned memory settings, missing indexes on custom fields, table bloat from neglected vacuuming, or inefficient code in custom modules; we measure with pg_stat_statements and query plans before changing anything.
Should we use managed PostgreSQL like Amazon RDS or run our own?
Managed PostgreSQL is usually the better choice when you lack in-house database operations skills, because patching, backups and failover are handled for you. Self-hosting gives full control over extensions, configuration and cost, and makes sense when someone is responsible for monitoring and restore testing.
How do you upgrade PostgreSQL to a new major version?
Major upgrades use either pg_upgrade, which is fast but needs a short downtime window, or logical replication to a new server, which keeps downtime to a brief switchover. Either way we rehearse on a copy of production and check extensions and application compatibility first.
Can PostgreSQL replace MongoDB for document data?
In many business applications, yes. JSONB columns store and index JSON documents inside PostgreSQL, so teams keep transactions and joins with relational data; MongoDB is still a better fit when the whole dataset is document-shaped and needs horizontal sharding.
What does a PostgreSQL backup strategy include?
A sound strategy includes full backups, continuous WAL archiving for point-in-time recovery, copies stored away from the database server, encryption, retention rules, and regular restore tests. A backup that has never been restored should not be trusted.
Next step
Is your Odoo or application database slowing you down?
Share your PostgreSQL version, server size and the symptoms you see. We'll review the configuration and slow queries and tell you what is worth fixing first.