Frequently asked questions

The questions that determine whether replication survives production.

Omni Loader supports initial load plus CDC, one-way and multi-master topologies, automatic failover, schema evolution, distributed agents, and mixed destinations.

Plan my migration Read documentation
Sources
What can it capture?

Oracle, SQL Server, PostgreSQL, MySQL, and Sybase.

Targets
Where can changes go?

Eleven database, stream, and Fabric sinks.

Topology
Where does it run?

On-prem, cloud, hybrid, relay-split, and HA.

Product and topology

Does it handle initial load, multi-master, and failover?

Yes. Seed + Stream catches up from a pinned position. Multi-master lets you choose the conflict winner per table and prevents loops automatically. HA adds leases, fencing, reassignment, and source-aware resume.

✓ Initial load begins a continuous pipeline
✓ One stream fans out to many sinks
✓ Choose the conflict winner per table
✓ Failover keeps one fenced owner

Operations and trust

What crosses the control plane?

Agents resolve secrets and move rows locally. Schema changes travel as ordered events. Durable spools absorb target and network outages.

✓ Row data stays in the data plane
✓ Outbound agents fit restrictive networks
✓ Unsafe DDL pauses visibly
✓ Delivery semantics are declared per path

Keep exploring

See the connected parts of the platform.

Documentation →

Move from answers to steps.

Sources →

Check the exact path.

Plan a pipeline →

Review your topology with us.

Ask the environment-specific question.

Tell us engines, network, write topology, tables, and recovery objectives. We will answer precisely.

Plan my migration Read documentation