PolyGres

Getting started

• Try it in a minuteThe docker run command and one-line client commands per protocol — the fastest way to see PolyWire actually working.
• Connecting to your PostgresWorked examples and real, previously-hit gotchas for pointing PolyWire at an existing on-prem Postgres, Supabase, Amazon RDS, Google Cloud SQL, Azure Database for PostgreSQL, or Oracle Cloud Infrastructure's Database with PostgreSQL.

PolyWire

Full capabilities → — every section below, and the on-page jump list to all of them, live on one page.

• OutcomesThe concrete result PolyWire is built to produce — a simpler architecture, a smaller attack surface, elastic scale, lower cost, and real operational visibility.
• Use casesGoverned AI access, gateway consolidation, connection pooling, sharding, distributed transactions, caching, rollups, standardizing on one database, a permanent compatibility shim for code not worth touching, and a mid-migration bridge — migration is one use case among several, not the whole story.
• ArchitectureThe shared eight-stage pipeline every protocol feeds — firewall, router, QoS, dialect translation, rollups, cache, stats — before reaching real Postgres.
• Admin consoleLive traffic metrics per protocol, per-tool MCP call counts, cache-hit-vs-Postgres RTT, and a sidebar covering monitoring, security, traffic, and configuration in one place.
• PolyCache benchmarksReal measured cache-hit vs. Postgres-round-trip numbers per protocol, with the full methodology.
• ShardingRouting by a real bound value decoded from each protocol's own wire format, with genuine cross-shard merge — not SQL-text pattern matching.
• SQSEnqueue vs. dequeue performance and how sqswire's JSON-over-HTTP framing compares to the binary wire protocols.
• RollupsPre-aggregated tables PolyWire creates and keeps fresh automatically, with transparent query rewriting.
• ObservabilityExporting metrics to CloudWatch, Datadog, New Relic, Grafana, Azure Monitor, GCP Cloud Monitoring, and AppDynamics — which platforms need a Collector in between and which take a direct push.
• Multi-AZ deploymentCross-zone cache backup replication, cluster discovery, and TLS between cache nodes.
• SecuritySQL firewall, connection ACL, QoS admission control, authentication (Postgres roles, per-user credentials, admin SSO), TLS, secrets management, and row-level security — each scoped honestly, including what's opt-in or narrower than the feature name suggests.
• Planned/unplanned outagesA scheduled drain waits for a clean cutover; unplanned failover redirects automatically — across every wire protocol.
• Error translationPer-protocol native error shapes (Oracle ORA-numbers, MySQL errno, SQL Server number/severity/state, MongoDB code/codeName, DynamoDB/SQS/OpenSearch native envelopes), a canonical-condition comparison table across all seven protocols, and the connection-loss SQLSTATE detail.

PolyAdvisor

• Full detailTwo ways to get an assessment (live connection or uploaded performance report), what you get (score, feature inventory, sizing, downloadable report), connecting and exploring with an LLM's help, and the admin console.

Setup reference

Source for both tools is private; the files below are the public documentation and image surface published alongside this site.

• README.mdTop-level repository reference.
• polywire/README.mdPolyWire-specific setup reference — full port list, environment variables, and build notes.
• polyadvisor/README.mdPolyAdvisor-specific setup reference.