Eventstream Doubt

[souvik666]

Suppose I use an EvenStream and I have 2 data sources with 2 different types of schemas, and I want to dump the data to the Evenhouse DB in two different tables. How do I segregate the 2 types of data from the EV_prod EvenStream so that 2 diff table land on 2 tables on the same EHDB?

[v-pnaroju-msft]

Thankyou, @rizalard0684 and @ati_puri for your responses.

Hi @souvik666,

We are following up to see if what we shared solved your issue. If you need more support, please reach out to the Microsoft Fabric community.

Thank you.

[ati_puri]

Hi @souvik666 ,

Excellenet article read from our Microsoft Fabric Community for the issue raised above. Handling Multiple Schemas in Eventstream for Azure... - Microsoft Fabric Community.

Please check if its useful for you.

Thanks

Ati Puri

[rizalard0684]

hey @souvik666 the answer by @arabalca  is correct and I'm aligned with him.

I just want to clarify some terminologies (item vs routing vs stream) because it's quite confusing:

1. Fabric EventStream item -> this is the Fabric item "EventStream" we create at Fabric workspace lavel
2. Inside Fabric EventStream item, this is (what I personally called) an EventStream routing -> this is your EV_prod
3. Inside Fabric EventStream routing, there is a stream -> this is your EV_prod-stream

Now comes my opinions:

1. Eventstream item has a single unified schema by default. All events flowing through it are merged into one routing / stream and makes it look like you can’t separate different source streams cleanly.
2. However, Fabric supports content-based routing inside one Eventstream item using operations (e.g. filter, manage fields and multiple Eventhouse destinations). https://learn.microsoft.com/en-us/fabric/real-time-intelligence/event-streams/route-events-based-on-...

In short, you can’t split into multiple eventsream routings/streams, however you can split into multiple Eventhouse tables but must do it via filters inside one Eventstream item. Hope this makes sense.

Appreciate if you can 'Kudos' and/or 'Accept as Solution' if this answered your query.

[v-pnaroju-msft]

Thankyou, @arabalca for your response.

Hi @souvik666,

We appreciate your inquiry through the Microsoft Fabric Community Forum.

We would like to inquire whether have you got the chance to check the solution provided by @arabalca to resolve the issue. We hope the information provided helps to clear the query. Should you have any further queries, kindly feel free to contact the Microsoft Fabric community.

Thank you.

[arabalca]

Hi @souvik666 ,

Your diagram already shows the correct approach! The pattern you've implemented is exactly what Microsoft recommends for this scenario. What is the questions?

What you've done right:

• Two separate sources (YellowTaxi and Kafka_Src) feeding into a single Eventstream (EV_prod)
• A Filter node per branch to separate the data by origin
• Two independent Eventhouse destinations (EH_Kafka and EH_Taxi) pointing to different tables in the same KQL database
• This is the standard way to route data from multiple sources with different schemas into separate Eventhouse tables — no code required, just the visual editor.

Official documentation:

Overview & multiple destinations: https://learn.microsoft.com/en-us/fabric/real-time-intelligence/event-streams/overview

If my comments helped solve your question, it would be great if you could like the two comments and mark it as the accepted solution. It helps others with the same issue and also motivates me to keep contributing.

Thanks a lot, I really appreciate it.