rabbitmq-diagnostics

Diagnose RabbitMQ node and cluster health

TLDR

Check that the local node is running

$ rabbitmq-diagnostics check_running

Print cluster status

$ rabbitmq-diagnostics cluster_status

Check for resource alarms (memory, disk)

$ rabbitmq-diagnostics check_alarms

Memory usage breakdown

$ rabbitmq-diagnostics memory_breakdown

Check listener ports are reachable

$ rabbitmq-diagnostics check_port_connectivity

Check virtual hosts are operational

$ rabbitmq-diagnostics check_virtual_hosts

Output as JSON for parsing

$ rabbitmq-diagnostics status --formatter json

SYNOPSIS

rabbitmq-diagnostics [options] command [commandoptions_]

rabbitmq-diagnostics provides health checks and observability commands for RabbitMQ nodes and clusters. It is the recommended tool for monitoring scripts and Kubernetes probes, replacing many ad-hoc uses of rabbitmqctl.Health checks return non-zero exit codes when a check fails, making them easy to wire into liveness/readiness probes and CI gates. The "deeper" checks (e.g. check_alarms) take longer and exercise more of the node, so use check_running for high-frequency probes and reserve heavier checks for less-frequent monitoring.

PARAMETERS

-n node

Target node name (default: rabbit@$(hostname)).

-l, --longnames

Use fully-qualified node names (FQDN).

-t timeout

Operation timeout, in seconds.

--formatter format

Output format: table (default), json, csv, erlang.

-q, --quiet

Suppress informational headers and status lines.

-s, --silent

No output for successful checks (useful in monitoring).

HEALTH CHECK COMMANDS

check_running

Node is running and the rabbit application is started.

check_local_alarms

No resource alarms in effect on the local node.

check_alarms

No resource alarms anywhere in the cluster.

check_port_connectivity

All listener ports accept TCP connections.

check_port_listener port

A specific port has an active listener.

check_protocol_listener proto

A specific protocol (amqp, mqtt, stomp) has an active listener.

check_virtual_hosts

All virtual hosts are running on the target node.

check_node_is_mirror_sync_critical

No classic-mirrored queues are in a sync-critical state.

OBSERVABILITY COMMANDS

status

High-level node status (versions, listeners, memory).

cluster_status

Membership, partitions, and per-node listener info.

memory_breakdown

Detailed memory accounting (queues, binaries, ETS, processes...).

environment

Effective application environment for all RabbitMQ apps.

listeners

List all configured network listeners.

runtime_thread_stats

BEAM scheduler/thread statistics.

CAVEATS

Requires a running RabbitMQ node and an Erlang cookie that matches the node's. Some checks need the management plugin or specific RabbitMQ versions. --formatter json is preferred for machine consumption — table output is not stable.

HISTORY

Introduced in RabbitMQ 3.8 (2019) as a dedicated diagnostics CLI, taking over health-check responsibilities that previously lived in rabbitmqctl.