PgDog vs SPQR
PgDog
pgdog.devSPQR
docs.pg-sharding.techSide by side
| Figure | PgDog | SPQR |
|---|---|---|
| Price from | Free | Free |
| Free plan | Yes | Yes |
| Free trial | — | — |
| Open source | — | — |
| Windows | Yes | — |
| macOS | Yes | — |
| Linux | Yes | Yes |
| API | — | Yes |
| Self-hosted | Yes | Yes |
| Plans published | 2 | 1 |
| Spec score · Database Proxy Software | 7.7 (#12) | 7.8 (#9) |
| Deployment | PgDog can be deployed with Kubernetes Helm, Docker, AWS ECS Terraform, pre-built binaries, or from source. | The getting-started guide provides a Docker image and lists router, admin console, and gRPC API ports as 6432, 7432, and 7000. |
| Integrations | The enterprise insights page lists Prometheus and OTEL exporters, plus Incident.io integration for alerts. | The FAQ says SPQR works with managed PostgreSQL databases including Yandex Managed PostgreSQL and AWS RDS, and names pgx and JDBC as compatible driver examples. |
| License | The project is free and open source under AGPL-3.0, and the documentation says internal use does not require sharing source code. | The source is distributed under the PostgreSQL Global Development Group License. |
| Security | The enterprise control plane page describes SSO for access protection, while installation documentation says user credentials can be encrypted at rest in supported environments such as Kubernetes or AWS Secrets Manager. | The docs show PostgreSQL connections configured with SSL modes including require and verify-full. |
| What it does | PgDog is a PostgreSQL connection pooler, load balancer, and database sharder designed to scale databases without application changes. | SPQR is a stateless PostgreSQL query router for horizontally scaling PostgreSQL through sharding. |
Plans
In their own words
PgDog
PgDog is a PostgreSQL connection pooler, load balancer, and database sharder designed to scale databases without application changes. It multiplexes many application connections across fewer PostgreSQL server connections and supports SET, LISTEN/NOTIFY, and advisory locks. Read queries can be distributed across replicas while writes go to the primary, with health checks and failover detection. Sharding keys guide query routing; queries without a key can run across databases in parallel. Cross-shard transactions use PostgreSQL prepared transactions and two-phase commit, with automatic rollback on errors. PgDog can copy data and stream row changes to another database without pausing query traffic. Deployment options include Kubernetes Helm, Docker, AWS ECS Terraform, pre-built binaries, and source. Releases include Linux arm64 and amd64, Apple Silicon Mac, and Debian packages; the macOS binary is unsigned. The project is free and open source under AGPL-3.0, and documentation says internal use does not require sharing source code. Enterprise support includes architecture reviews, a dedicated Slack channel, and an SLA promising a response within 15 minutes.
SPQR
SPQR is a self-hosted query router for horizontally scaling PostgreSQL through sharding. It can extract a sharding key from a query and route it to the relevant key range and shard when possible; keys can use one column or several. Queries can also specify a key or shard in a SQL comment. Applications connect using the PostgreSQL protocol, and the router and coordinator provide administrative consoles accessible through psql. SPQR supports one-phase and two-phase commit strategies; two-phase commit provides atomicity across shards and requires max_prepared_transactions greater than zero on each shard. The system can move data between shards to balance workload, and the project describes reducing transferred ranges to limit locking impact. Its documentation includes Docker and configuration-file deployment instructions. SPQR is free and distributed under the PostgreSQL Global Development Group License. The FAQ cautions that it is not suited to complex cross-shard queries, transactions modifying multiple shards or OLAP workloads, and reports a roughly 10–30% performance drop versus vanilla PostgreSQL depending on workload.