Cluster Compatibility

Supported stacks

Component	Version	Notes
Aeron Java/C media driver	1.53.3	`aeron` submodule tag; archive + consensus module
Aeron Cluster Java	1.53.3	clustered service container
rusteron (Rust bindings)	0.2.13 (workspace `Cargo.toml`)	`rusteron-client`, `rusteron-media-driver`; bundles Aeron C 1.53.3 (the tutorial sample also uses `rusteron-archive`)
ergo-sbe (codec generator)	0.1.x	schema 111 session + mark codecs
Rust	1.99+	the workspace `rust-version` (the cluster `Cargo.toml` declares none); edition 2024. CI builds the cluster crate with 1.99.0; the MSRV job checks `ergo-sbe` only
OS	Linux x86-64 and arm64	CI runs the cluster unit tests on `ubuntu-24.04` and `ubuntu-24.04-arm`, and the live-cluster harness on `ubuntu-24.04` only. macOS is not run in CI
Java	25 (17+)	Archive and Cluster; the test harness runs Temurin 25 in CI (21 for the Aeron 1.46.8 compatibility job). Java for Archive and Cluster

Schema identities

Schema	Id	SBE `version`	`semanticVersion`
Aeron Cluster Session	111	17	5.4
Aeron Cluster Mark	110	2	5.4

These are pinned in cluster/schemas/ and generated by cluster/build.rs. The generated codec API is unstable — use the high-level AeronCluster client, not the cluster_codec_types module directly.

Failure modes

The client handles these documented transitions:

Trigger	Expected behaviour
Leader loss (the session’s egress image closes / publication CLOSED)	Session → `AwaitingNewLeader`, poll for `NewLeaderEvent`
`NewLeaderEvent` (the last one in a poll)	Bind to the image that delivered it; add the leader’s ingress publication without waiting; `Connected` once it connects, `Disconnected` if it has not within the message timeout
`SessionEvent(CLOSED)`	Session → `PendingClose` → `Closed`, even while a new leader’s publication is pending
Redirect during connect	Follow `leader_endpoints` to new leader
New ingress publication (redirect or new leader)	Keep every unicast ingress channel setting but the endpoint, and re-add IPC and multicast channels unchanged, so the message limits do not change
Auth challenge	Pass to `CredentialsSupplier`, send `ChallengeResponse`
Auth rejection	`ClusterError::AuthRejected`
Timeout (connect / poll / new-leader)	`ClusterError::Timeout` / `ClusterError::Disconnected`
Listener panic	`ClusterError::ListenerPanicked`, session continues
Malformed egress frame	`ClusterError::ProtocolError`
Payload too large	`ClusterError::PayloadTooLarge`
Publication backpressure	Retryable `ClusterError::Publication`

What CI runs against a live cluster

The harness (cluster-test-harness/tests/) runs in ci.yml (aeron-cluster-integration, Aeron 1.53.3, Java 25) and in cluster-compatibility.yml (Aeron 1.46.8, Java 21, when cluster/ or sbe/src/ changes).

Covered	Test file
3-node connect, leader failover, redirect	`failover.rs`, `failover_own_driver.rs`, `client_parity.rs`, `nonblocking.rs`
New leader that never connects times out	`nonblocking.rs` (`a_new_leader_that_never_connects_times_out`)
`SessionEvent(CLOSED)` while a new leader is pending	`nonblocking.rs`
Ingress channel settings kept across failover and redirect	`ingress_channel_settings.rs`
Fragmented egress, regular and controlled poll	`egress_fragmentation.rs`
Restart, reconnect, quorum loss, log recovery	`restart.rs`
Null and simple credentials (the default authenticator accepts all)	`auth.rs`

Not covered by the harness: auth challenge and rejection, controlled-poll abort, and connect or poll timeout expiry. Unit tests in cluster/src/ (run by ci.yml’s test job) cover challenge dispatch (egress.rs), the challenge response (credentials.rs, client.rs) and the new-leader timeout (client.rs, test_new_leader_timeout_disconnects). No test drives ClusterError::AuthRejected, a controlled-poll abort, or a connect or poll timeout.

Limitations

No dynamic membership (static cluster only).
No snapshot/recovery integration.
No TLS/mTLS at the Aeron transport layer.
The cluster_codec_types module is #[doc(hidden)] and not stable API.