Skip to main content

Provider configuration

pocketPG can detect your cloud provider automatically from the connection hostname, or you can set it explicitly in the provider configuration modal.

Supported providers

ProviderAuto-detectionCloud metricsTuning awareness
Aiven*.aivencloud.comCPU, memory, disk via Aiven APIavn service update CLI instead of ALTER SYSTEM SET
Amazon RDS*.rds.amazonaws.comPlannedManaged-provider checkpoint caveat
Amazon Aurora*.rds.amazonaws.comPlannedManaged-AWS behavior with Aurora-specific notes
Azure*.postgres.database.azure.comPlannedManaged-provider checkpoint caveat
Google Cloud SQL*.cloudsql.google.comPlannedManaged-provider checkpoint caveat
Supabase*.supabase.coNot integratedManaged-provider caveat and direct-connection guidance
Neon*.neon.techNot integratedManaged-provider caveat and direct-connection guidance
Self-managedNo detection neededN/AFull ALTER SYSTEM SET support

What provider configuration enables

Cloud metrics

When provider credentials are configured, pocketPG can pull CPU, memory, and disk metrics from the provider API. Aiven is currently the most complete integration.

Supabase and Neon are handled differently. They do not need extra provider credentials in pocketPG today. pocketPG uses the PostgreSQL connection itself for product monitoring, and stores the provider identity only so the UI can show the right setup guidance and managed-service caveats.

Instance shape

Cloud provider APIs can expose instance shape details that make recommendations more realistic than PostgreSQL settings alone.

Provider-aware tuning

On Aiven, global PostgreSQL configuration changes must go through Aiven service configuration instead of ALTER SYSTEM SET. pocketPG detects this and shows provider-appropriate copy commands where possible. Parameters that the provider manages internally are marked as service-managed.

On Supabase and Neon, the main value of the provider profile today is simpler onboarding:

  • pocketPG can preselect the provider from the hostname
  • setup guidance can tell you to use the direct connection instead of a pooler endpoint
  • managed-service checkpoint notes stay attached to the instance
  • future provider-specific monitoring can be added without changing the saved instance identity
info

Per-table ALTER TABLE DDL (fillfactor, autovacuum scale factors) works on all providers, including Aiven. Only global ALTER SYSTEM SET is restricted on managed services.

Managed-provider checkpoint caveat

On managed PostgreSQL services, checkpoints_req can increment from provider-internal housekeeping (backups, archiving, replication). pocketPG's checkpoint card appends a note on managed providers advising you to verify the WAL rate before acting on checkpoint-frequency recommendations.

This applies to Supabase and Neon too. They do not need a separate monitoring tab to be useful in pocketPG, but they do benefit from being classified correctly as managed PostgreSQL.

See also: