AWS Adds Shared EventBridge Bus for Multi-Account Event Systems

AWS has added an enhanced custom event bus to Amazon EventBridge, giving large organizations a way to run one shared event backbone across accounts instead of stitching together multiple buses with cross-account rules.

The release targets a scaling problem inside event-driven applications.

A single custom event bus can work while one team owns the architecture, but enterprise adoption often pushes teams into separate AWS accounts.

Routing events between those accounts can then require bus-to-bus configurations, repeated permission work and charges that grow with each hop.

The new EventBridge option changes that architecture by making a centralized bus shareable across an AWS organization.

Publishers can send events to the shared bus, while consuming teams subscribe only to the events they need.

Platform teams keep a broader view of event flows and control over who can publish or consume without provisioning separate routing paths for every team.

AWS is also using the launch to separate the new resource from the existing EventBridge model.

Existing custom event buses continue to operate as Custom event bus – classic, while the enhanced custom event bus appears as the recommended new option in the EventBridge console.

Adoption is optional, so current workloads do not have to move before teams are ready.

The technical tradeoff is clearest in ordering.

Many event-driven systems are designed so event order does not matter, but some workflows, such as logistics applications using driver-location updates, need messages to arrive in sequence.

The enhanced bus lets publishers include an EventGroupId, then delivers matching events in order to subscribers that choose ordered delivery.

Other subscribers on the same bus can still consume the events asynchronously.

AWS tied that ordered processing to synchronous invocation for targets such as AWS Lambda.

In that mode, EventBridge confirms that processing succeeded before acknowledging the event, reducing the need to place Amazon SQS between a bus and Lambda solely to add reliability for ordered work.

The release also introduces a Subscriber resource that gives each consumer one place to define matching logic, destination settings, retry behavior and failure handling.

That replaces a setup in which teams configure rules, targets and retry behavior across several EventBridge resources.

For platform teams, the simpler resource model is meant to make each consumer’s event requirements easier to see and manage.

Deduplication is another part of the new bus design.

Publishers can turn on content-based deduplication so EventBridge identifies duplicate retries from the event payload and drops matching events that arrive within five minutes.

AWS describes that option as useful for sources that cannot reliably attach their own retry identifier, while applications that already use idempotency tokens can keep doing so.

The pricing model also shifts for multi-account deployments.

Costs are split between ingestion on the publishing side and delivery on the consuming side, replacing a per-event pattern in which cross-account and bus-to-bus routing costs can compound.

AWS says the enhanced custom event bus is available now in selected US, European and Asia Pacific regions, including Northern Virginia, Ohio, Oregon, Ireland, Frankfurt, Stockholm, Spain, Hong Kong, Malaysia, Mumbai, Singapore, Sydney, Thailand and Tokyo.