Roadmap

PgDog is being actively developed. Along with performance improvements and bug fixes, new features are constantly being added. The goal for PgDog is to manage most aspects of at-scale Postgres production operations, and we're just getting started.

Chart

Stable and in production.

Partially built, but what's already there is working.

On the roadmap, but not started yet.

Not currently planned.

〃 Shorthand for "ditto" or "same as above".

Foundational features

These features are required for PgDog to act as a replacement for PgBouncer and/or PgCat.

Feature	Status	Notes
Transactional pooler
Session mode		No sharding or load balancing.
Load balancer
Health checks & failover
Prepared statements
Metrics		Admin database views contain more columns than PgBouncer.
Encryption
Authentication		Password authentication only. `scram-sha-256`, `md5` are supported.

Sharding

These features are required for PgDog to shard Postgres databases without application changes.

Query engine

Query engine provides a uniform view over multiple shards. Clients can use regular SQL statements to read and write data, as if they were using a normal (unsharded) database.

Feature	Status	Notes
Direct-to-shard reads		Sharding key must be specified in the query.
Direct-to-shard writes		Sharding key must be specified in the query. Multi-tuple `INSERT`s not supported yet.
Cross-shard queries		Partial aggregates and sorting support. CTEs & subqueries not supported yet.
Cross-shard CTEs		#380
Cross-shard subqueries		#381
Cross-shard joins		#94
Cross-shard transactions		Supports two-phase commit. Not benchmarked yet.
Omnisharded tables		Unsharded tables with identical data on all shards.
Rewrite queries		Alter queries to support aggregate/sorting by rows not returned in result set.
`COPY`		Sharding key must be specified in the statement and the data. Supports text, CSV, and binary formats only.
Multi-statement queries		e.g.: `SELECT 1; SELECT 2;`. First query is used for routing only, entire request sent to the same shard(s). #395.

Aggregates

Support for aggregate functions in cross-shard queries.

Aggregate function	Status	Notes
`SUM`		Target column(s) must be present in returned rows.
`COUNT`		〃
`MIN`		〃
`MAX`		〃
`AVG`		#434
Percentile distributions		Could be expensive to calculate, need spill to disk.

Sorting

Support for sorting rows in cross-shard queries.

Sorting function	Status	Notes
`ASC column`		Target column(s) must be present in returned rows.
`DESC column`		〃
`ASC a, b, c`		Sorting by multiple columns works. Columns must be present in returned rows.
`DESC a, b, c`		〃
`<->`		`pgvector` sort by L2. Both columns must be in result set.

Resharding

Resharding splits existing database(s) into more shards, without requiring downtime.

Feature	Status	Notes
Data sync		Sync table data with logical replication. Not benchmarked yet.
Schema sync		Sync table, index and constraint definitions. Not benchmarked yet.
Online rebalancing		Not automated yet, requires manual orchestration.

Schema & data integrity

Manage table schema(s) and ensure data constraints are enforced across multiple shards.

Feature	Status	Notes
Primary keys		`BIGINT` and `UUID` partially supported for hash-based sharding only. #386. Other data types require cross-shard unique index support.
Unique indexes		Enforce uniqueness constraints across an unsharded column(s). #439.
`CHECK` constraints		They are generally arbitrary SQL checks and need to be executed prior to row updates.
Schema validator		Check that all shards have identical tables, indexes, etc.

Orchestration

Features around production deployments.

Feature	Status	Notes
DNS cache		Override DNS TTL to accelerate failover. Built but not documented.
Primary failover to standby		Move writes from a failed primary database to a standby replica, promoting it. Not currently planned.
Configuration sync		Synchronize config changes between instances of PgDog. Admin command to pause traffic exists, but no automation.