[Questions] quorum queue create filed

[dormanze]

Community Support Policy

• I have read RabbitMQ's Community Support Policy
• I run RabbitMQ 4.x, the only series currently covered by community support
• I promise to provide all relevant information (versions, logs from all nodes, rabbitmq-diagnostics output, detailed reproduction steps)

RabbitMQ version used

4.0.6

Erlang version used

26.2.x

Operating system (distribution) used

linux

How is RabbitMQ deployed?

Generic binary package

rabbitmq-diagnostics status output

See https://www.rabbitmq.com/docs/cli to learn how to use rabbitmq-diagnostics

# PASTE OUTPUT HERE, BETWEEN BACKTICKS

Logs from node 1 (with sensitive values edited out)

See https://www.rabbitmq.com/docs/logging to learn how to collect logs

# PASTE LOG HERE, BETWEEN BACKTICKS

Logs from node 2 (if applicable, with sensitive values edited out)

See https://www.rabbitmq.com/docs/logging to learn how to collect logs

# PASTE LOG HERE, BETWEEN BACKTICKS

Logs from node 3 (if applicable, with sensitive values edited out)

See https://www.rabbitmq.com/docs/logging to learn how to collect logs

# PASTE LOG HERE, BETWEEN BACKTICKS

rabbitmq.conf

See https://www.rabbitmq.com/docs/configure#config-location to learn how to find rabbitmq.conf file location

# PASTE rabbitmq.conf HERE, BETWEEN BACKTICKS

Steps to deploy RabbitMQ cluster

config rabbitmq and start

Steps to reproduce the behavior in question

1、create a classic queue
2、delete classic queue
3、create quorum queue with same name

advanced.config

See https://www.rabbitmq.com/docs/configure#config-location to learn how to find advanced.config file location

# PASTE advanced.config HERE, BETWEEN BACKTICKS

Application code

# PASTE CODE HERE, BETWEEN BACKTICKS

Kubernetes deployment file

# Relevant parts of K8S deployment that demonstrate how RabbitMQ is deployed
# PASTE YAML HERE, BETWEEN BACKTICKS

What problem are you trying to solve?

We want to change classic queue to quorum queue(without changing the queue's name).
The sample code for switching is as follows:

package changequeue;

import com.rabbitmq.client.*;

import java.io.IOException;
import java.util.HashMap;
import java.util.Map;
import java.util.concurrent.TimeoutException;

public class ChangeQueueTypeToQuorum {
    public static void main(String[] args) throws IOException, TimeoutException {
        ConnectionFactory factory = RabbitmqFactory.getFactory("/");
        Connection baseConnection = factory.newConnection( "oldConnection");
        Map map = new HashMap<>();
        map.put("x-message-ttl", 1800000);
        Channel baseChannel = baseConnection.createChannel();
        String queueName = "test-queue";
        // ======================================================
        // create classic queue for test
        baseChannel.queueDelete(queueName);
        baseChannel.queueDeclare(queueName, true, false, false, map);
        // ======================================================

        // The actual process of switching queues
        Connection connection = factory.newConnection("newConnection");
        final Channel channel = connection.createChannel();
        Channel newChannel;
        try {
            newChannel = channel.getConnection().createChannel();
        } catch (IOException e) {
            throw new RuntimeException(e);
        }
        try {
            Map quorumMap = new HashMap<>();
            quorumMap.put("x-message-ttl", 1800000);
            quorumMap.put("x-queue-type", "quorum");
            newChannel.queueDeclare(queueName, true, false, false, quorumMap);
        } catch (IOException e) {
            try {
                newChannel = newChannel.getConnection().createChannel();
                newChannel.queueDelete(queueName);
                newChannel.queueDeclare(queueName, true, false, false, quorumMap);
            } catch (IOException ex) {
                throw new RuntimeException(e);
            }
        }  finally {
            try {
                newChannel.close();
            } catch (IOException e) {
                throw new RuntimeException(e);
            } catch (TimeoutException e) {
                throw new RuntimeException(e);
            }
        }
        try {
            channel.queueBind(queueName, "amq.topic", "test");
            channel.basicConsume(queueName, true, "consumer",new DefaultConsumer(channel) {
                @Override
                public void handleDelivery(String consumerTag, Envelope envelope, AMQP.BasicProperties properties, byte[] body) {
                }
            });
        } catch (IOException ex) {
            throw new RuntimeException(ex);
        }
    }
}

However, I have noticed that some queues fail to be created and no matter how many times I retry, they cannot be successfully created. The only way to recover is by restarting the RabbitMQ server. The related error messages are as follows:

2025-08-29 05:51:29.074640+08:00 [error] <0.8453354.0> Channel error on connection <0.8453199.0> (7.225.3.47:49250 -> 7.227.23.15:5673, vhost: '/', user: 'rabbitmq_ma'), channel 2:
2025-08-29 05:51:29.074640+08:00 [error] <0.8453354.0> operation queue.declare caused a channel exception not_found: no queue 'ucmarket-update-job.instance-update' in vhost '/'
2025-08-29 05:51:29.096379+08:00 [warning] <0.8453354.0> Channel '<0.8453354.0>' will be closed.
2025-08-29 05:51:29.096515+08:00 [warning] <0.8453354.0> Some consumers : '#{}' created by channel : '<0.8453354.0>' are die out.
2025-08-29 05:51:29.097253+08:00 [info] <0.8453199.0> Channel '<0.8453371.0>' will be created by connection [<<"7.225.3.47:49250 -> 7.227.23.15:5673">>] user [<<"rabbitmq_ma">>] in vhost [<<"/">>] .
2025-08-29 05:51:31.878995+08:00 [info] <0.8453206.0> queue 'ucmarket-update-job.instance-update' in vhost '/': delivery_limit not set, defaulting to 20
2025-08-29 05:51:31.879141+08:00 [info] <0.8453206.0> queue 'ucmarket-update-job.instance-update' in vhost '/': delivery_limit not set, defaulting to 20
2025-08-29 05:51:31.879313+08:00 [info] <0.8453206.0> queue 'ucmarket-update-job.instance-update' in vhost '/': delivery_limit not set, defaulting to 20
2025-08-29 05:51:31.879676+08:00 [error] <0.8453588.0>     supervisor: {<0.8453588.0>,ra_server_sup}
2025-08-29 05:51:31.879676+08:00 [error] <0.8453588.0>     errorContext: start_error
2025-08-29 05:51:31.879676+08:00 [error] <0.8453588.0>     reason: {already_started,<0.2324155.0>}
2025-08-29 05:51:31.879676+08:00 [error] <0.8453588.0>     offender: [{pid,undefined},
2025-08-29 05:51:31.879676+08:00 [error] <0.8453588.0>                {id,'%2F_ucmarket-update-job.instance-update'},
2025-08-29 05:51:31.879676+08:00 [error] <0.8453588.0>                {restart_type,transient},
2025-08-29 05:51:31.879676+08:00 [error] <0.8453588.0>                {significant,false},
2025-08-29 05:51:31.879676+08:00 [error] <0.8453588.0>                {shutdown,5000},
2025-08-29 05:51:31.879676+08:00 [error] <0.8453588.0>                {child_type,worker}]
2025-08-29 05:51:31.879676+08:00 [error] <0.8453588.0>
2025-08-29 05:51:31.879943+08:00 [error] <0.8453578.0> ra: failed to start a server #{id => {'%2F_ucmarket-update-job.instance-update','rabbit@rabbitmqservice-1'},machine => {module,rabbit_fifo,#{name => '%2F_ucmarket-update-job.instance-update',max_length => undefined,max_bytes => undefined,queue_resource => {resource,<<47>>,queue,<<117,99,109,97,114,107,101,116,45,117,112,100,97,116,101,45,106,111,98,46,105,110,115,116,97,110,99,101,45,117,112,100,97,116,101>>},created => 1756417891879,dead_letter_handler => undefined,become_leader_handler => {rabbit_quorum_queue,become_leader,[{resource,<<47>>,queue,<<117,99,109,97,114,107,101,116,45,117,112,100,97,116,101,45,106,111,98,46,105,110,115,116,97,110,99,101,45,117,112,100,97,116,101>>}]},overflow_strategy => drop_head,delivery_limit => 20,expires => undefined,msg_ttl => undefined,single_active_consumer_on => false}},membership => voter,friendly_name => [113,117,101,117,101,32,39,117,99,109,97,114,107,101,116,45,117,112,100,97,116,101,45,106,111,98,46,105,110,115,116,97,110,99,101,45,117,112,100,97,116,101,39,32,105,110,32,118,104,111,115,116,32,39,47,39],cluster_name => '%2F_ucmarket-update-job.instance-update',uid => <<50,70,95,85,67,77,52,71,68,84,74,67,50,69,80,56,57,89>>,initial_members => [{'%2F_ucmarket-update-job.instance-update','rabbit@rabbitmqservice-1'},{'%2F_ucmarket-update-job.instance-update','rabbit@rabbitmqservice-2'},{'%2F_ucmarket-update-job.instance-update','rabbit@rabbitmqservice-0'}],log_init_args => #{max_checkpoints => 3,min_checkpoint_interval => 8192,snapshot_interval => 8192,uid => <<50,70,95,85,67,77,52,71,68,84,74,67,50,69,80,56,57,89>>},metrics_key => {resource,<<47>>,queue,<<117,99,109,97,114,107,101,116,45,117,112,100,97,116,101,45,106,111,98,46,105,110,115,116,97,110,99,101,45,117,112,100,97,116,101>>},ra_event_formatter => {rabbit_quorum_queue,format_ra_event,[{resource,<<47>>,queue,<<117,99,109,97,114,107,101,116,45,117,112,100,97,116,101,45,106,111,98,46,105,110,115,116,97,110,99,101,45,117,112,100,97,116,101>>}]},tick_timeout => 5000}, error: {error,
2025-08-29 05:51:31.879943+08:00 [error] <0.8453578.0>                                                                                                                                                                                                                                                                                                                                                                        

client receive error message:

create queue with UI:

but creat classic queue with same name success.

[kjnilsson]

It looks like you are either declaring the same queue names in parallel or you have some remnants of a prior failed declaration. Deletes can sometimes fail or timeout which occasionally could leave partial queue "clusters" behind. Make sure that your delete operation in your code succeeds before proceeding with the rest.

[dormanze]

Thank you for your response.
The current issue is that creation consistently fails, even when I stop all other creation and deletion actions; creation still fails on the UI.

[michaelklishin]

@dormanze you were told what to do: remove the queue first, e.g. with rabbitmqctl delete_queue, which can handle several potential failure scenarios and force delete the queue record.

RabbitMQ 4.0.x is out of community support, so without a repeatable way to reproduce against 4.1.x or main, that's all we can suggest.