Queries related to HA in Quorum queues

76 views
Skip to first unread message

Abhishek Nayak

unread,
Nov 12, 2025, 6:25:11 AMNov 12
to rabbitmq-users

Hi Team ,

We have  a 2 node cluster with RMQ 3.13.1 and Erlang 26.2.2 .

We have by default, the vhost has quorum queues and around 5k pending messages in total all queues

Today , when we reboot one node , we see below behaviour :

1. Unable  to publish message to cluster , on to a queue whose leader node is down  . Ideally , cluster should take message and synchronize to both the nodes once all the nodes are up

2. We saw some message loss during the node restart , Is it general behaviour for a quorum queue to loose messages when node is restarted ?

3. How should we ensure High availability of messages  / persistence even in case of node down in this version of RMQ and Quorum queues.

Thanks,

Abhishek

 

Vilius Šumskas

unread,
Nov 12, 2025, 6:32:58 AMNov 12
to rabbitm...@googlegroups.com

Hi,

 

start from using at least 3 nodes for your cluster https://www.rabbitmq.com/docs/clustering#node-count and see how it goes.

 

--

   Best Regards,

    Vilius

--
You received this message because you are subscribed to the Google Groups "rabbitmq-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to rabbitmq-user...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/rabbitmq-users/e63b225a-4fd5-41ad-ba95-2f5769198509n%40googlegroups.com.

Abhishek Nayak (Consultant)

unread,
Nov 12, 2025, 6:34:29 AMNov 12
to rabbitmq-users, Lakshmi Sravani Kappagantula (Consultant), Brahmarao Mothe (Consultant)
++ Lakshmi and Brahmi in CC

From: 'Abhishek Nayak' via rabbitmq-users <rabbitm...@googlegroups.com>
Sent: Wednesday, November 12, 2025 3:25 AM

To: rabbitmq-users <rabbitm...@googlegroups.com>
Subject: [rabbitmq-users] Queries related to HA in Quorum queues
 

WARNING: This email originated from outside of Qualcomm. Please be wary of any links or attachments, and do not enable macros.

--
You received this message because you are subscribed to a topic in the Google Groups "rabbitmq-users" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/rabbitmq-users/ap53voqGOPU/unsubscribe.
To unsubscribe from this group and all its topics, send an email to rabbitmq-user...@googlegroups.com.

Abhishek Nayak

unread,
Nov 12, 2025, 6:42:05 AMNov 12
to rabbitmq-users
Hi Vilius,

You mean to say , if we are using Quorum queues in our cluster , that it should be always of 3 node cluster ??

What is the issue , if we have 2 node cluster ?? Ideally if the messages get replicated to other node with Quorum queues  , then there should not be any message failures at least during node restarts.

Could you please advise on the same.

Thanks,
Abhishek 

Michal Kuratczyk

unread,
Nov 12, 2025, 7:01:06 AMNov 12
to rabbitm...@googlegroups.com
It's already covered in the docs
additionally: https://www.rabbitmq.com/docs/quorum-queues#quorum-requirements

if you want more details - read about Raft protocol (it's not RabbitMQ specific).




--
Michal
RabbitMQ Team
Reply all
Reply to author
Forward
0 new messages