Goal: Be able to run, manage and observe a single service mesh running across multiple OpenShift clusters

Problem: Istio users may run a service mesh over multiple OpenShift clusters, as of now, they can’t have a single mesh that span multiple clusters.

Why is this important: For scalability and redundancy, OpenShift users need to spread their workload across multiple clusters, those scenarios aren’t currently supported.

Dependencies (internal and external):

• Istio: Most of the work has been done already

Definitions:

• Clustered: a set of Service Mesh control planes which act as a single control plane and are configured as a single entity
• Federated: a set of Service Mesh control planes which interact with each other and are configured independently

Prioritized epics + deliverables (in scope / not in scope):

• Investigate the mechanics of a clustered solution supporting replicated components and configuration
• Investigate necessary operator functionality to support clustering configurations
• Investigate the mechanics of a federated solution supporting interacting Service Mesh control planes
• Investigate necessary operator functionality to support federating configurations

Estimate (XS, S, M, L, XL, XXL):

• XL for Maistra

Previous Work:

• Work from Istio

OSSM Team Use Case / Design Artifacts:

• Maistra Federation Use Cases & Design Research 
• OSSM Federation Sync Meeting Notes (Weekly Meeting - Tuesdays @ 10am EST)
• Maistra Federation Design Document
• Diagram: ServiceImport and ServiceExport act like ServiceEntries in a Federated Mesh