<!-- Generated from the rendered page by scripts/write-llm-mirrors.mjs. Do not edit by hand. -->
Canonical: https://molo17.com/blog/gluesync-supabase-target-agent/
Markdown mirror: https://molo17.com/blog/gluesync-supabase-target-agent/index.md
Title: New Supabase target agent in Gluesync 2.3 · MOLO17
Description: New Supabase target agent in Gluesync 2.3: direct connection or session pooler, TLS by default, row level security, pgvector and PostGIS types written.

[← All articles](/blog/)

Article

# New in Gluesync 2.3: the Supabase target agent

![New in Gluesync 2.3: the Supabase target agent featured image](/blog/gluesync-supabase-target-agent/0c56bf57-gluesync-supabase-target-1x1.png)

16 Nov 2026  [Gluesync](/blog/?category=gluesync)  5 min read

Gluesync 2.3 adds a new Supabase target agent. Snapshots and CDC changes land in Supabase Postgres tables over the PostgreSQL wire protocol, with the PostgreSQL target's type mapping and bulk COPY path and TLS on by default.

Gluesync 2.3 adds a NEW **Supabase target agent**. It writes snapshots and CDC changes into Supabase Postgres tables over the PostgreSQL wire protocol, with TLS on by default. Supabase is PostgreSQL, so the agent is the Gluesync PostgreSQL target with Supabase's connection defaults: the same type mapping, the same snapshot and CDC write paths, the same bulk `COPY` path.

Supabase is a target in Gluesync. Any supported source feeds it through a pipeline in Core Hub: the snapshot first, then the source agent's CDC stream. Gluesync needs only a database connection and does not call the Supabase REST API or management API, so it works with projects hosted on supabase.com and with self-hosted Supabase.

Docs: [Supabase agent](https://docs.molo17.com/gluesync/latest/agents/supabase-intro.html) · [Target setup](https://docs.molo17.com/gluesync/latest/agents/supabase-target.html) · [Changelog](https://docs.molo17.com/gluesync/latest/agents/supabase-agent-changelog.html).

## Why a dedicated Supabase agent

Supabase speaks plain PostgreSQL, yet a project has its own shape: database `postgres`, a `postgres` role that is not a superuser, encrypted connections, a pooler with a session mode and a transaction mode, and a set of schemas Supabase manages for itself. The agent starts from those defaults (database `postgres`, user `postgres`, TLS on) and knows what to leave alone:

-   The schemas Supabase manages (`auth`, `storage`, `realtime`, `vault`, `extensions`, `graphql` and the like) are treated as system schemas by the full-catalog walk of Core Hub: they are not walked or sampled, so `auth.users` and its password hashes are never read
-   Your own tables, in `public` or in schemas you created, are walked as usual
-   The `postgres` role is enough: the agent creates its `GLUESYNC` helper schema and its staging tables as that role

## When to use it

When an application built on Supabase needs data that lives in other databases, and you want it in Supabase Postgres tables with the changes following continuously. Tables you create yourself, for example in a Supabase migration, are written as they are, and Gluesync can also create the target tables as it does for any SQL target.

## What the agent does

-   Writes snapshots and CDC changes into Supabase Postgres tables through the bundled PostgreSQL JDBC driver
-   Uses the PostgreSQL target's bulk path: rows are staged in a table of the `GLUESYNC` schema with `COPY`, then merged into the target table, for snapshots and for CDC
-   Supports connection pooling and Core Hub's _TRUNCATE before snapshot_ option
-   Writes tables with row level security enabled, also forced
-   Writes extension types such as `vector` (pgvector), PostGIS `geometry`, `geography` and `raster`, and `hstore`

## How it connects

| Connection | Supported | Notes |
| --- | --- | --- |
| Direct connection, `db.<project-ref>.supabase.co:5432` | Yes | Recommended. IPv6 only unless the project has the IPv4 add-on. User `postgres` |
| Session pooler, `aws-0-<region>.pooler.supabase.com:5432` | Yes | IPv4 and IPv6. User `postgres.<project-ref>` |
| Transaction pooler, port `6543` | No | The write path uses prepared statements, which the transaction pooler does not support |

A Core Hub without IPv6 connectivity can use the session pooler, which has IPv4 addresses.

### TLS

TLS is on by default. With _Trust server certificate_ on (the default) the JDBC `sslmode` is `require`: the connection is encrypted and the certificate is not checked against a CA. With it off, `sslmode` is `verify-ca`: import the project's CA certificate, downloadable from the Supabase dashboard, into the trust store you give the agent.

### Row level security

The `postgres` role bypasses row level security on Supabase, so a table needs no policy for Gluesync, also with `FORCE ROW LEVEL SECURITY`. A connection user without `BYPASSRLS` follows the table's policies and needs policies that let it insert, update, delete and select the rows.

## Data types

The mapping is the PostgreSQL target's. The integration test suite writes these types as values and as primary keys, through the snapshot and CDC paths, and compares them with the source:

| Group | Types |
| --- | --- |
| Integers and booleans | `smallint`, `integer`, `bigint`, `oid`, `boolean`, serial types |
| Numbers | `numeric`, `decimal`, `money`, `real`, `double precision` (NaN and infinities included) |
| Dates and times | `date`, `time`, `timetz`, `timestamp`, `timestamptz`, `interval`, dates before 1582 included |
| Text | `varchar`, `char`, `text`, `json`, `jsonb`, `xml`, `uuid`, enums, `tsvector`, `tsquery` |
| Binary and bits | `bytea`, `bit(n)`, `bit varying` |
| Network | `inet`, `cidr`, `macaddr`, `macaddr8` |
| Geometric and ranges | `point` to `polygon`, all range and multirange types |
| Arrays | Arrays of the types above |
| Extension types | `vector`, `geometry`, `geography`, `raster`, `hstore` |

A `numeric` value of `NaN` cannot be read or written: the Gluesync decimal type has no such value.

## Updates with bulk operations during CDC

With bulk operations on, an update during CDC is written as a delete of the existing row and an insert of the new one. On Supabase that has visible effects worth knowing before you turn it on:

-   `UPDATE` triggers do not fire, while `DELETE` and `INSERT` triggers do, so a `moddatetime` trigger on `updated_at` does not run
-   Foreign keys with `ON DELETE CASCADE` delete their child rows on every update, `ON DELETE SET NULL` clears them, and `RESTRICT` or `NO ACTION` makes the write fail while child rows exist
-   Supabase Realtime subscribers and anything reading the change log see a `DELETE` and an `INSERT` instead of an `UPDATE`

Entities whose target table has an identity column always apply updates as `UPDATE` statements.

## Connection setup

From _Connect_ in the Supabase dashboard, take the direct connection or the session pooler details. In Core Hub, add Supabase as the **target** of a pipeline:

-   **Hostname**: `db.<project-ref>.supabase.co` or `aws-0-<region>.pooler.supabase.com`
-   **Port**: `5432`
-   **Database name**: `postgres`
-   **Username**: `postgres` on the direct connection, `postgres.<project-ref>` on the session pooler
-   **Password**, **Enable TLS** (on by default) and **Trust server certificate** (on by default)

The same setup through the Core Hub REST API:

```bash
curl --location --request PUT 'http://core-hub-ip-address/pipelines/{pipelineId}/agents/{agentId}/config/credentials' \
--header 'Content-Type: application/json' \
--header 'Authorization: ••••••' \
--data '{
  "hostCredentials": {
    "connectionName": "myAgentNickName",
    "host": "db.<project-ref>.supabase.co",
    "port": 5432,
    "databaseName": "postgres",
    "username": "postgres",
    "password": "",
    "maxConnectionsCount": 100,
    "enableTls": true,
    "trustServerCertificate": true
  }
}'
```

Supported versions: Supabase Postgres 15 or later, hosted or self-hosted. The integration test suite runs against Supabase Postgres 17.6.

## Where it fits

Bringing data from the databases a business already runs into the Postgres behind a Supabase application, kept current with CDC from the sources in the Gluesync catalog. The full list of sources and targets is on the [Gluesync](https://molo17.com/gluesync/) page.

## Next steps

1.  Read the [Supabase target setup](https://docs.molo17.com/gluesync/latest/agents/supabase-target.html).
2.  Deploy **Gluesync 2.3.0.3** and add Supabase as the target of a pipeline in Core Hub.
3.  Point it at your project's direct connection or session pooler and watch the tables fill.

For a step-by-step setup, follow [Set up a Supabase target](https://docs.molo17.com/gluesync/latest/learn/how-to-set-up-a-supabase-target.html).

Want to see it on your project? [Request a free trial](https://molo17.com/get-gluesync/) or [book a demo](https://molo17.com/book-a-demo/).

## Gluesync integrations

-   Gluesync for Supabase
    
    -   [Replicate to Supabase →](/solutions/replicate-to-supabase/)
-   Gluesync for PostgreSQL
    
    -   [CDC from PostgreSQL →](/solutions/postgresql-cdc/)
    -   [Replicate to PostgreSQL →](/solutions/replicate-to-postgresql/)

[← Older article Install Gluesync on Windows and Linux: the lean path to the Control Plane](/blog/install-gluesync-on-windows-and-linux/)

## Keep reading

1.  [Gluesync Courier: Gluesync pipelines now speak REST in both directions Courier brings REST APIs into Gluesync as sources and targets: describe an API once in Core Hub, then send changes out or take records in with the same pipelines. 12 Nov 2026](/blog/gluesync-courier-http-source-target-rest-api/)
2.  [Gluesync Gluesync 2.3.2: point-in-time restart, disaster recovery, lake file queries and Podman Gluesync 2.3.2: replay CDC from any moment, run a standby Core Hub site, query lake files in Query Studio, deploy on Podman and let a speed test tune your snapshots. 11 Nov 2026](/blog/gluesync-2-3-2-0-point-in-time-restart-disaster-recovery-lake-queries-podman/)
3.  [Gluesync New in Gluesync 2.3: the Microsoft Fabric Open Mirroring target agent Gluesync 2.3 adds a new Microsoft Fabric Open Mirroring target agent. Snapshots and CDC changes land as Parquet files with row markers in the landing zone of a Fabric mirrored database, and Fabric keeps the Delta tables up to date. 9 Nov 2026](/blog/gluesync-microsoft-fabric-open-mirroring-target-agent/)

[Back to all articles](/blog/) [More in Gluesync →](/blog/?category=gluesync)
