Solution · Federated data access

Query distributed databases as one SQL system

Run read-only SQL across isolated Gluesync agents through one JDBC endpoint—without first copying raw data into a central warehouse.

ONE ENDPOINT · MANY DATABASES SQL CLIENT · DBEAVER read-only SELECT c.name, o.total FROM pg.customers c JOIN mongo.orders o jdbc:gluesync://corehub Query Forge Apache Calcite planner filter pushdown projection pushdown no data copy PostgreSQL agent-eu-west read-only MongoDB agent-catalog read-only Db2 for i agent-ibmi read-only
  1. 01 Use one JDBC endpoint
  2. 02 Resolve a federated query
  3. 03 Combine read-only results

The challenge

Some questions span systems

Operational data remains distributed across engines, networks, and ownership boundaries even when teams need one answer.

01 / Silos

Cross-system analysis becomes ETL

A simple join can trigger another extraction pipeline, staging area, and copy of sensitive operational data.

02 / Access

Clients need many credentials

Each database connection adds secrets, network routes, drivers, and governance work for every consumer.

03 / Control

Ad-hoc queries can affect production

Unbounded access risks expensive scans, writes, and resource contention on operational databases.

How it works

One gateway, distributed execution

Query Forge exposes Gluesync-connected databases through a branded JDBC endpoint and executes work close to each agent.

  1. 01

    Connect

    Use one JDBC endpoint

    BI tools and SQL clients connect through jdbc:gluesync instead of opening a route to every database.

  2. 02

    Plan

    Resolve a federated query

    The gateway identifies participating agents and pushes eligible filters and projections toward each source.

  3. 03

    Return

    Combine read-only results

    Query Forge joins the result sets and returns one SQL response without creating a central raw-data copy.

Gluesync capabilities

Federation with access boundaries intact

Query Forge combines familiar SQL connectivity with scoped authentication and safe execution defaults.

01

Cross-agent SQL

Join relational and supported non-relational systems behind different agents from one query.

  • Apache Calcite planning
  • Filter and projection pushdown
  • Heterogeneous schema access
02

JDBC for existing tools

Connect familiar clients and BI applications through the Gluesync JDBC driver.

  • jdbc:gluesync endpoint
  • DBeaver and DataGrip workflows
  • Tableau-compatible connectivity
03

Role-aware access

Authenticate with local credentials or Personal API Tokens, with access bounded by the user's permissions.

  • Read-only execution
  • Scoped PAT access
  • User-bound authorization

Ask cross-system questions without building another copy

Explore Query Forge or talk with MOLO17 about federated access across your connected database estate.