RabbitMQ

RabbitMQ hosting with any node count, in any cloud

RabbitMQ as a service on dedicated clusters in OSStend's cloud or your cloud account, and RabbitMQ on premise in your data center, run 24/7 on AceMQ LTS, AceMQ's long-term support build.

45 minutes with an engineer: we review your workload and sketch the design.

  • Operated by AceMQ: 160+ customers, 10+ years
  • 24/7 on-call engineers
console.osstend.com / acme-corp / rabbitmq-orders

An example design file, with illustrative values:

service:    rabbitmq-orders
engine:     rabbitmq
version:    3.13-acemq-lts     # pinned, monthly CVE patches
nodes:      5                  # sized to measured load
zones:      [eu-west-1a, 1b, 1c]
plugins:
  - shovel
  - federation
  - delayed_message_exchange
federation:
  upstream: dc2.example.internal   # on premise
upgrade_window: "Sun 02:00-04:00 CET, after change freeze"
region:     customer-aws-account/eu-west-1
on_call:    24/7, named engineers
osstend.yaml (example)Design signed off

RabbitMQ hosting from OSStend is for teams whose RabbitMQ no longer fits a plan: clusters sized to the workload, federation and shovel to your data center, the plugins your workload needs and upgrades on your change calendar. Engineers run it 24/7, in a cloud or on premise.

160+AceMQ customers
10+Years running these systems
24/7Engineers on call
3Places it runs: our cloud, yours, your data center
Engines we run, on the platforms you already use
  • PostgreSQL
  • MySQL
  • MongoDB
  • Redis
  • Valkey
  • OpenSearch
  • Kafka
  • RabbitMQ
  • AWS
  • Google Cloud
  • Azure
  • Kubernetes
  • OpenShift
  • Prometheus
  • Grafana
  • Datadog
  • Dynatrace
  • Splunk
Built for your workload

What OSStend tailors for RabbitMQ

RabbitMQ is where tailoring matters most. A cluster is designed around the node count, the plugins, the links between sites and the version your applications are validated on, and the design is signed off before anything is built.

OSStend cloud3 zonesrabbitmq-orders
  • Metrics to your Prometheus or Datadog
  • AceMQ engineers on call 24/7
Example design, illustrative values: the real one is sized from your measured load.
AreaSet for your workload
Size and shapeNode count, instance size and storage chosen from measured load, including five-node clusters. Replicated queues are quorum queues, which are Raft-based and confirm an operation once a majority of members agree, with a default group of three, so queue count and message size feed the sizing as much as message rate.
TopologyMulti-zone clusters, and federation and shovel between sites, including links to brokers in your data center. The shovel plugin moves messages one way from a source to a destination, in the same cluster or another, and federation is built for links that fail from time to time. Which of the two each link uses is decided per flow.
Configs and pluginsBroker settings and the plugins the workload depends on, such as shovel, federation and the delayed-message exchange, with full configuration and operator access for OSStend engineers. Dynamic shovels change at runtime, while static shovels need a node restart.
Versions and upgradesAceMQ LTS, AceMQ's long-term support build, with monthly CVE patching and security fixes backported. Upgrades follow your change calendar and are never forced. The version matters to queue design: classic queue mirroring was removed in RabbitMQ 4.0, so mirrored queues are replaced by quorum queues before a cluster crosses that line.
Where it runsOSStend's cloud, your cloud account, or RabbitMQ on premise in your data center, on Kubernetes, OpenShift, Windows hosts or air-gapped sites, with FIPS where required.
MonitoringMetrics and alerts forwarded to your Datadog, Dynatrace, Splunk or Prometheus. On-call engineers watch queue depth, unacked messages, connection and channel counts, and the state of each quorum queue group.

There is no fixed menu here: the node count, plugins, settings and upgrade dates your workload needs are written into the design you sign off before anything is built, and that design is what OSStend runs.

Moving off mirrored classic queues is the RabbitMQ task that most often forces a design decision. The RabbitMQ guide says there is no in-place conversion from a classic queue to a quorum queue, because the two differ at the storage level. The routes are a blue-green deployment, or a new virtual host on the same cluster where quorum queues are declared under the old names and federation moves the traffic across. Policies that still carry ha-mode and related keys are cleaned up on the way.

Built for your workload

Where turnkey RabbitMQ services stop

Amazon MQ for RabbitMQ documents limits on cluster size, settings, plugins, upgrades and connections. They belong to the plan menu, not to RabbitMQ, and each row below sets one beside the design OSStend would write.

Area and managed serviceDocumented limitWhat OSStend does insteadSource
Cluster sizeAmazon MQ for RabbitMQA cluster is three nodes, one per availability zone, or a single instance.What OSStend does insteadClusters sized to your measured load, including five-node clusters.AWS docs, checked 17 Nov 2026 ↗
SettingsAmazon MQ for RabbitMQAWS lists five broker values you can change: consumer timeout, heartbeat, operator policy changes, relaxed quorum queue redeclaration checks and HTTP security headers.What OSStend does insteadFull configuration and operator access for OSStend engineers, who change settings on request.AWS docs, checked 30 Oct 2026 ↗
PluginsAmazon MQ for RabbitMQAWS documents a list of 11 supported plugins. Static shovels are not supported, a shovel cannot target a private broker, federation cannot use an upstream in a private broker, and the delayed-message exchange and stream plugins are not on the list.What OSStend does insteadThe plugins the workload needs, and federation or shovel to any site.AWS docs, checked 17 Nov 2026 ↗
UpgradesAmazon MQ for RabbitMQAWS upgrades brokers to the latest patch version in the maintenance window, and moves a broker to the next minor version when its current minor version reaches end of support.What OSStend does insteadAceMQ LTS builds with monthly CVE patching, and upgrades on your change calendar.AWS docs, checked 17 Nov 2026 ↗
ConnectionsAmazon MQ for RabbitMQPer-node connection, channel and queue ceilings cannot be overridden, for example 5,000 connections on an mq.m7g.large.What OSStend does insteadLimits set by the cluster design, not by an instance tier.AWS docs, checked 17 Nov 2026 ↗
One cloudAmazon MQAmazon MQ runs on AWS only.What OSStend does insteadThe same service in AWS, Azure, Google Cloud and your data center, on one contract.AWS docs, checked 30 Oct 2026 ↗

The Amazon MQ alternative page sets these limits beside the OSStend design, with the path off Amazon MQ, and the managed service limits index lists every documented limit with its source and the date checked.

Compare

RabbitMQ hosting plan limits: CloudAMQP and Amazon MQ

A RabbitMQ hosting plan fixes the node count, the connection ceiling and the price tier. CloudAMQP's dedicated plans offer one, three or five nodes, and Amazon MQ offers one instance or three nodes. Figures below were read from each vendor's page on 17 Nov 2026.

Illustrative: plan names and shapes stand in for any provider's menu.
OSStend
Node countsOne, three or five nodes; the largest plan, Raging Rhino, is listed for one node only (CloudAMQP plans)A single instance, or a cluster of three nodes (AWS docs)Any node count, sized to the measured load
Entry price by shape$50 a month on one node (Sassy Squirrel, 1.5k connections); from $297 on three nodes (Big Bunny, 9k connections); from $1,495 on five nodes (Roaring Rabbit, 20k connections) (CloudAMQP plans)Billed hourly by instance size; a three-node cluster uses three instances and costs three times one (AWS pricing)Quoted per design; see pricing
Connection ceilingPer plan: 1.5k (Sassy Squirrel) to 80k (Loud Lion) on one node, and up to 200k on a five-node Loud Lion (CloudAMQP plans)Configurable only up to a maximum per instance type: 5,000 on mq.m7g.large up to 160,000 on mq.m7g.16xlarge for a single instance, and per node in a cluster, where the real limit can be lower (AWS docs)Set by the cluster design, not by a plan tier
Shared and free tiersLittle Lemur is free (20 connections, 100 queues); Tough Tiger is $19 a month (100 connections, 1,000 queues) (CloudAMQP plans)New AWS accounts get Free Tier credits of up to $200 (AWS pricing)No free tier and no card sign-up
PluginsNot stated on the plans pageA documented list of 11 plugins; static shovels are not supported (AWS docs)The plugins the workload needs, including federation and shovel to any site
Billing and switchingBilled by the second; plans can be switched without downtime; data transfer is included up to 30% of team spend (CloudAMQP plans)Instance hours, storage and data transfer (AWS pricing)A price written into the contract after the design review
Where it runsAWS, Google Cloud, Azure, DigitalOcean and Scaleway (CloudAMQP plans)AWS onlyOSStend's cloud, your cloud account or your data center

Sources: CloudAMQP plans and pricing page, AWS Amazon MQ pricing page and developer guide, linked in each cell; read 17 Nov 2026. OSStend does not publish plan prices: each cluster is priced from its design.

Plans suit a RabbitMQ whose queues and connections sit inside one tier. When the cluster needs a node count between the listed shapes, a plugin the vendor does not list, federation to a data center or an upgrade on your own calendar, that is the case for a designed cluster. If a plan covers your numbers, use the plan.

How it works

Design. Move. Run. for RabbitMQ

For RabbitMQ, an engineer designs the nodes, queue types and links between sites and you sign it off, OSStend carries the messages across and owns the cutover, and named engineers then run the cluster 24/7.

  1. Design

    An engineer reviews your workload (message rates, queue types, connections, failure modes, compliance and where it must run) and writes the cluster design: node count, topology, plugins and version. You sign it off before anything is built.

  2. Move

    OSStend runs the migration and owns the cutover. Federation, shovel and proxies carry messages between the old and the new cluster while both sides stay live, and applications move when you are ready. Classic mirrored queues are re-declared as quorum queues on the way, since they cannot be converted in place. Zero downtime is the plan.

  3. Run

    Named engineers run it 24/7: upgrades on your calendar, backups, scaling, plugin and configuration changes on request, quorum queue health and incident response, with an SLA written into your contract.

Deployment

Deployment and service levels

RabbitMQ runs dedicated in OSStend's cloud or inside your own cloud account, both fully managed. RabbitMQ on premise runs co-managed in your data center, including Windows hosts, FIPS and air-gapped sites.

AceMQ engineersOne on-call team, one contract, wherever it runs
  • Dedicated in OSStend's cloud

    Single-tenant RabbitMQ clusters on Google Cloud or AWS, with nodes spread across zones so a quorum can survive losing one.

    Fully managed
  • In your cloud account

    RabbitMQ inside your cloud account and VPC, so message data, the node bill and the network boundary stay yours. AWS first, then Azure and Google Cloud.

    Fully managed
  • In your data center

    RabbitMQ on premise: Windows hosts, Kubernetes, OpenShift or your own servers, with FIPS where required and air-gapped sites covered, run through an operator.

    Co-managed

The fully managed options carry an uptime commitment, and RabbitMQ on premise carries a 24/7 response-time commitment because the servers are yours. The service levels page sets out the tiers, and deployment shows each option in detail.

Uptime and response figures are set per plan and written into your contract. AceMQ, which operates OSStend, sets its support response targets from 15 minutes, by plan.

Operated by AceMQ

Proof from AceMQ's operating record

The upgrade, queue and cluster work for RabbitMQ is done by AceMQ's engineers. AceMQ operates OSStend with 160+ customers, 10+ years of running open-source infrastructure and engineers on call 24/7, as acemq.com states.

Operated by AceMQCompany figures, as published on acemq.com
160+AceMQ customers
10+years operating
24/7on-call engineers
Insurance · named reference

IBM MQ to RabbitMQ, without replacing the core

American National Insurance moved its Windows-based IBM MQ queues to RabbitMQ on Kubernetes first, with both brokers live until behaviour was validated, while IBM MQ licensing costs were progressively removed. Use case on acemq.com.

AceMQ engagement
Insurance · named reference

A Kafka CDC pipeline it did not need

American National replaced a Kafka-based change data capture pipeline for SQL Server with standalone Debezium publishing to RabbitMQ, removing Kafka clusters and Kafka Connect the use case did not need. Use case on acemq.com.

AceMQ engagement

Other customers stay unnamed unless AceMQ holds their written permission. A reference with a similar queue topology can be raised in the 45-minute review.

Pricing

Pricing for RabbitMQ hosting

RabbitMQ hosting is priced from four drivers: the node count and queue design, the storage and retention, the SLA tier and where it runs. An engineer prices the design after a workload review, and there is no free tier or card sign-up for RabbitMQ.

Node count and queue design
three nodes or five, and how many quorum queue groups they carry, set how much compute and disk the cluster needs.
Cluster size
node count, instance size and storage, which follow message rates and retention.
SLA tier
the uptime commitment on a fully managed cluster, or the 24/7 response-time commitment on RabbitMQ on premise.
Where it runs
OSStend's cloud, your cloud account or your data center, which decides who pays for the nodes.

Share the message rates, queue types, connection counts and where RabbitMQ must run, and the engineer prices that cluster, not a plan size.

The pricing page shows how the model applies across every engine.

Enterprise only

When to contact OSStend, and when not to

Contact OSStend when RabbitMQ needs a node count, plugin set, link between sites or version timing that a broker plan will not give you. If a plan covers the queues, use it.

Contact us if

  • Your RabbitMQ needs five or more nodes, or settings and plugins a plan does not allow.
  • You need federation or shovel between cloud and data center, or RabbitMQ on premise.
  • You must hold a validated version through a change freeze.
  • You are leaving Amazon MQ or another RabbitMQ service and want the migration run for you.
Talk to an engineer

Not a fit if

  • A standard broker plan carries your queues, and it costs less than a designed cluster.
  • You want a free tier, or to sign up with a card. For RabbitMQ, OSStend begins with a conversation with an engineer.
  • You will keep running RabbitMQ yourself and only want a support contract: speak to AceMQ.
FAQ

Questions about RabbitMQ hosting

These are the questions platform and integration teams ask most before moving RabbitMQ to a hosted service.

Talk to an engineer →

What is RabbitMQ hosting, and how is it different from RabbitMQ as a service?

RabbitMQ hosting runs the broker cluster for you on infrastructure you do not manage day to day, and RabbitMQ as a service is the same idea sold as a subscription. OSStend adds a design step first, so node count, plugins, version and upgrade timing are set for the workload instead of picked from a plan.

Can RabbitMQ run on premise and still link to a cloud cluster?

Yes. RabbitMQ on premise runs co-managed in your data center, on Kubernetes, OpenShift, Windows hosts or air-gapped sites. Federation and shovel link it to clusters in OSStend's cloud or your cloud account, so messages flow between sites.

What happened to mirrored classic queues?

Classic queue mirroring was removed in RabbitMQ 4.0 after a long deprecation, and non-replicated classic queues still work. Replicated queues are now quorum queues or streams. A classic queue cannot be converted in place, so the move is a blue-green deployment or a new virtual host with federation carrying traffic, and a cluster of at least three nodes is the minimum for high availability.

How does a move from Amazon MQ avoid downtime?

Federation, shovel and proxies carry messages between the old and the new cluster while both sides stay live. Applications move in stages, and traffic switches only after the new side is verified. The Amazon MQ alternative page describes the path, and the Amazon MQ to RabbitMQ migration guide covers the phases, cutover and rollback.

What is AceMQ LTS?

AceMQ LTS is AceMQ's long-term support build, compatible with RabbitMQ®, with monthly CVE patching and security fixes backported. It lets a cluster stay on a validated version without forced upgrades.

How many nodes can a RabbitMQ hosting plan have?

It depends on the vendor. CloudAMQP's plans page lists one, three or five nodes for dedicated RabbitMQ, and Amazon MQ offers a single instance or a three-node cluster (both read 17 Nov 2026). OSStend sizes the node count to the measured load, including five-node clusters.

What does RabbitMQ hosting cost?

The price follows the node count, queue design, the SLA tier and where it runs. The pricing page sets out the model, and an engineer prices the specific design after a workload review.

Talk to an engineer

Talk to an engineer about your RabbitMQ cluster

45 minutes with an engineer: bring your message rates, your queue types and the version you run, and the engineer sketches the cluster design.

  • A design sketch for your workload, not a plan menu
  • A zero-downtime cutover path for what you run today
  • A price for that design, after the review
See pricing →

Your request goes to the OSStend team. No newsletter sign-up.

Get a quoteTalk to an engineer
2Engines
3Where should it run?
4What do you need?

Your request goes to the OSStend team. No newsletter sign-up.