Microsoft Fabric has widened the ways it can receive events from platforms many companies already run. With connectors for Cribl, Solace PubSub+ and SAP Datasphere, Eventstream can bring those streams into Real-Time Intelligence and route them to destinations inside Fabric.
The change matters for teams that keep messaging, observability or SAP replication outside Fabric and want to analyze the data in near real time without building a custom integration for every source. A connector, however, does not remove the work of defining the schema, credentials, transformations, destination and day-to-day operation of the stream.
What each connection solves
- Cribl: connects the data-routing platform to Eventstream. It is a way to take events you already collect and forward them to destinations in Fabric.
- Solace PubSub+: receives messages from the Solace messaging platform so they can be processed and routed in the stream.
- SAP Datasphere: uses a Datasphere replication flow to send data to a dedicated Eventstream source.
After ingestion, Eventstream can apply transformations and route events to supported destinations such as an Eventhouse, a Lakehouse or a derived stream. That makes it possible to combine operational telemetry, business events and replicated data in near-real-time analysis.
An example: tracking orders without waiting for the nightly load
Picture an operation that publishes order changes to Solace and replicates master data from SAP Datasphere. Eventstream can receive the events, filter out invalid messages and send the stream to an Eventhouse. A team can then query recent delays and failures, and the consolidated data can also feed a semantic model for Power BI reporting.
This design does not mean Power BI replaces the streaming layer. Eventstream and Eventhouse handle event ingestion and querying; the semantic model organizes measures and dimensions for business analysis. Keeping those responsibilities separate avoids trying to use report refresh as a low-latency mechanism.
What to validate before production
- Access and network: validate credentials and connectivity between the source and the Fabric workspace. Requirements vary with the connector and the source configuration.
- Schema and evolution: check names, data types and optional fields. Payload changes can break transformations or downstream queries.
- Volume and retention: estimate events per second, average size and retention time. Test behavior under peaks, not just the average.
- Duplicates and ordering: decide how to handle resent and out-of-order events; both affect counts and KPIs.
- Monitoring and recovery: track failures, lag and capacity consumption. Define how to restart the stream and reconcile periods of downtime.
Availability and requirements for each connector can change. Confirm the status shown in the Fabric documentation for your organization's region and tenant before planning a migration or committing to production.
How to get started
In a workspace with Fabric capacity, create an Eventstream and add the matching source. Configure the connection following the specific documentation for Cribl, Solace PubSub+ or SAP Datasphere, validate a sample of the stream, and only then add transformations and destinations. Run a proof of concept with representative data and clear latency and quality criteria first.
These connectors reduce the friction of putting existing event platforms on Fabric's analytics path. The real gain comes when the integration is backed by schema governance, monitoring and a destination that fits the kind of queries you need.