Ambient multi-network multicluster support is now Beta

Source: Istio Blog

Our team of contributors has been busy throughout the transition to 2026. A lot of work was done to get the multi‑network multicluster for Ambient to a production‑ready state. Improvements were made in areas from our internal tests up to the most popular multi‑network multicluster asks in Ambient, with a big focus on telemetry.

Gaps in Telemetry

The benefits of a multicluster distributed system are not without their trade‑offs. Some complexity is inevitable at larger scale, making good telemetry even more important. The Istio team recognized several gaps that needed to be covered. Starting with release 1.29, telemetry is now more robust and complete when the Ambient data plane operates over distributed clusters and networks.

If you’ve deployed alpha multicluster capabilities before in multi‑network scenarios, you might have noticed some source or destination labels appear as “unknown”.

• In a local cluster (or clusters sharing the same network), Waypoint and Ztunnel are aware of all existing endpoints via xDS.
• In multi‑network deployments, replicating all endpoint information across separate networks makes xDS peer discovery impractical, resulting in missing peer information when requests cross network boundaries to reach a different Istio cluster.

Telemetry Enhancements

Istio 1.29 ships with augmented discovery mechanisms in its data plane for exchanging peer metadata between endpoints and gateways across different networks. The HBONE protocol is now enriched with baggage headers, allowing Waypoint and Ztunnel to exchange peer information transparently through east‑west gateways.

Peer Metadata Flow (L7 Metrics)

Diagram showing peer metadata exchange across different networks

1. The client in Cluster A initiates a request; Ztunnel establishes an HBONE connection through the Waypoint.
2. Ztunnel sends a CONNECT request with a baggage header containing downstream peer metadata. The Waypoint stores this metadata.
3. The baggage header is stripped, and the request is routed normally to a different cluster.
4. On the receiving side, the Ztunnel in Cluster B receives the HBONE request, replies with success, and appends a baggage header containing the upstream peer metadata.
5. The upstream metadata becomes visible to the Waypoint as the response passes back, enabling it to emit metrics about both parties.

Note: This functionality is behind a feature flag. To enable the telemetry enhancements, set the AMBIENT_ENABLE_BAGGAGE feature option.

Other Improvements and Fixes

• Connectivity: Ingress gateways and Waypoint proxies can now route requests directly to remote clusters, simplifying resiliency and enabling more flexible design patterns for multicluster and multi‑network deployments.
• Stability: Several smaller fixes improve the stability and robustness of multi‑network multicluster.
• Documentation: The multicluster documentation has been updated, including a new guide on setting up Kiali for an Ambient multi‑network deployment.

Limitations and Next Steps

• The work primarily targets multi‑network support; multicluster in single‑network deployments remains in alpha.
• The east‑west gateway may preferentially select a specific endpoint for a period of time, affecting load distribution for requests coming from a different network. This behavior impacts both Ambient and sidecar data‑plane modes, and we plan to address it for both cases.
• Ongoing collaboration with the Istio community aims to close the remaining gaps.

We’re excited to release this beta and look forward to your feedback. The future looks bright for Istio multi‑network multicluster.

If you would like to try out Ambient multi‑network multicluster, please follow the setup guide. Remember, this feature is in beta status and not ready for production use. We welcome bug reports, thoughts, comments, and use cases—reach us on GitHub or Slack.