Feasibility of Using Confluent Cloud Schema Registry with GCP Managed Kafka

Confluent Cloud
1

Hi Team,

I’m exploring the possibility of integrating Confluent Cloud Schema Registry with an external Kafka cluster, specifically GCP Managed Kafka (Google Cloud Managed Kafka / Apache Kafka service).

I went through the Confluent Cloud documentation but couldn’t find any clear indication that Schema Registry can be used with Kafka clusters that are not hosted within Confluent Cloud.

Could you please confirm whether Confluent Cloud Schema Registry supports integration with external Kafka clusters such as GCP Managed Kafka?
If not, are there any recommended alternatives or supported patterns for using Confluent Schema Registry in this scenario?

Any documentation, guidance, or official confirmation would be very helpful.

Thank you!

2

hey @Archita

it depends a bit on your use case though in general you should be able to access Confluent Cloud SR via public endpoint.

Not sure if this officially supported but I think it should work.

What about a self-hosted SR? it this feasible?

best,

Michael

3

Hi @mmuehlbeyer — thanks for the quick response!

To clarify our use case:

We are evaluating whether Confluent Cloud (Schema Registry) can act as the central registry for an external Kafka cluster — in our case, GCP Managed Kafka (Google Cloud’s fully managed Kafka service, not Confluent Cloud Kafka).

Our producers/consumers would run on GCP and connect to:

  • Kafka: GCP Managed Kafka

  • Schema Registry: Confluent Cloud SR

We want to confirm whether this cross-integration is officially supported, especially around:

  • Compatibility of Confluent serializers/deserializers with non-Confluent Kafka when the SR is in Confluent Cloud

  • Authentication/authorization patterns (API key/secret for SR only)

  • Any known limitations (e.g., subject naming strategies, compatibility checks, schema versioning, etc.)

Regarding self-hosting:
Yes, hosting our own Schema Registry is possible (Bitnami/Confluent images on Kubernetes), but before going that route we wanted to confirm whether Confluent Cloud SR can be used independently without requiring Confluent Cloud Kafka.

If you or someone from the Confluent team could confirm the official support stance or point us to documentation, that would really help us plan the architecture.

Thanks again!

4

hey @Archita

thanks for clarifying.

Might be something @Joe could help with, regarding official supported scenarios I’m a bit over-asked.#

Best,

Michael

5

Thanks so much, @mmuehlbeyer — really appreciate your help so far.

@Joe, if you don’t mind jumping in — we’d be grateful for your guidance here.

We’re evaluating whether Confluent Cloud Schema Registry can be reliably and officially used with an external Kafka cluster, specifically GCP Managed Kafka (Google Cloud’s managed Kafka offering). We want to ensure that this architecture is either directly supported or at least not discouraged from Confluent’s side.

It would be extremely helpful if you could clarify:

  • Whether Confluent Cloud SR is supported as a standalone component for non-Confluent Kafka clusters

  • Any limitations, caveats, or unsupported patterns we should be aware of

  • Recommended alternatives if this setup isn’t advisable

Your insights would really help us make an informed decision and move in the right direction.

Thanks in advance for your time!

6

hey @Archita

just thinking out loud:

  • is there a specific reason you’d like to stick with Googles cloud kafka or are you free to switch to Confluent as well?
  • I’ve seen that there is a beta offering for SR in Google cloud, is this something you could try?
  • what about self-hosting the SR?

best,

Michael

7

Hey Michael,

Thanks for checking in — happy to clarify.

Right now, we’re not strictly tied to Google’s Managed Kafka. We’re evaluating both options (GCP Managed Kafka vs Confluent) based on integration complexity, governance capabilities, cost, and long-term operational overhead. So switching to Confluent is still on the table if it gives us a cleaner and more feature-complete setup.

Regarding Google’s Schema Registry beta — we did take a look. It’s quite limited at the moment (reduced API coverage, and still missing some ecosystem integrations etc), so it doesn’t fully meet our needs today. But we’re keeping it as a fallback option if improvements roll out.

So in summary: all options are open, and we’re just trying to converge on whatever gives us the most stable and maintainable solution long-term.

Happy to discuss more if you have thoughts or experience with any of these paths.

8

hey @Archita

thanks for providing these insights.

so what I would probably do is, test 2 options:

  • Google Kafka with Confluent Cloud SR
  • Confluent Cloud Kafka and SR

to test this you should be able to easily create an account in Confluent Cloud

the following :backhand_index_pointing_down:

provides a promocode which will add 400$ credits to your account :slight_smile:

this should probably be enough to test everything

best,

michael