Integration glossary
Data hub
A data hub is a central store that connected systems feed and read, so they all share one current version of records such as customers, orders and assets.
Hub, warehouse and lake
A data hub is operational. It holds the current version of shared records, and changes flow in from the systems and back out to them while the business runs.
A data warehouse is analytical. It keeps history arranged for reporting, and it is usually loaded on a schedule and read rather than written back. A data lake stores raw data in bulk, often before anyone has decided how it will be used. A company can have all three, and a hub often feeds the other two.
Hub and spoke instead of point to point
Without a hub, systems connect to each other in pairs and the number of connections grows quickly with every new system. With a hub, each system connects once, to the centre, and reads what the others have written there.
What a hub needs to work
-
An agreed model of the shared records, so every system means the same thing by a customer.
-
Change capture, so the hub stays current without full reloads.
-
Conflict handling, for when two systems change the same record.
-
Lineage, so every value shows which system it came from.
The STRAX hub
At the centre of STRAX is a hub, a SQL Server database on your own servers that every connected system feeds and reads. By default it keeps a record after a source removes it, or, where you switch it on per table, it marks the record deleted and passes the delete on. The reports, portals and workflows you build in STRAX read the same hub, so everyone works from the same facts.
How the STRAX integration engine works → What a canonical data model is → What two-way synchronisation is → What soft deletes and delete propagation are →
More terms
Other ideas worth knowing.
MCP server
A service that offers a system's tools and data to AI assistants over an open protocol, within limits the system sets.
Read the definition →Soft delete and delete propagation
Marking a record deleted while keeping the row, and passing a delete on to the other systems that hold the same record.
Read the definition →Worker pool and horizontal scaling
A group of machines sharing one queue of work, so capacity grows by adding another machine to the pool.
Read the definition →See it on your own systems.
A demo takes about an hour and shows these ideas working on systems like yours.