Linux RabbitMQ-server cluster

https://www.rabbitmq.com/clustering.html
https://kamaok.org.ua/?p=2689
https://blog.flux7.com/blogs/tutorials/how-to-creating-highly-available-message-queues-using-rabbitmq

1) add hostname (do real hostname) between master and slave to /etc/hosts

2) copy /var/lib/rabbitmq/.erlang.cookie from master to slave

systemctl restart rabbitmq-server (on slave)


3) join to cluster

rabbitmqctl stop_app

Stopping rabbit application on node rabbit@<hostname_slave> ...

rabbitmqctl reset

Resetting node rabbit@<hostname_slave> ...

rabbitmqctl join_cluster rabbit@<hostname_slave>

Clustering node rabbit@<hostname_slave> with rabbit@<hostname_master>

rabbitmqctl start_app

Starting node rabbit@<hostname_slave> ...
completed with 0 plugins.


4) check cluster

rabbitmqctl cluster_status

Cluster status of node rabbit@<hostname_slave> ...

[{nodes,[{disc,[rabbit@<hostname_master>,rabbit@<hostname_slave>]}]},

{running_nodes,[rabbit@<hostname_master>,rabbit@<hostname_slave>]},

{cluster_name,<<"rabbit@localhost">>},
{partitions,[]},

{alarms,[{rabbit@<hostname_master>,[]},{rabbit@<hostname_slave>,[]}]}]

5) Configure syncing queues police between master and slave nodes

Run on master

rabbitmqctl set_policy ha-all "" '{"ha-mode":"all","ha-sync-mode":"automatic"}'

for all vhost

https://stackoverflow.com/questions/38611583/rabbitmq-is-there-a-way-to-apply-policy-on-all-virtual-host-in-cluster-environ

from
https://www.rabbitmq.com/ha.html

ha-sync-mode: manual
- this is the default mode. A new queue mirror will not receive existing messages, it will only receive new messages. The new queue mirror will become an exact replica of the master over time, once consumers have drained messages that only exist on the master. If the master queue fails before all unsychronised messages are drained, those messages will be lost. You can fully synchronise a queue manually, refer to unsynchronised mirrors section for details.

ha-sync-mode: automatic
- a queue will automatically synchronise when a new mirror joins. It is worth reiterating that queue synchronisation is a blocking operation. If queues are small, or you have a fast network between RabbitMQ nodes and the ha-sync-batch-size was optimised, this is a good choice.

for / vhost
rabbitmqctl set_policy ha-all "" '{"ha-mode":"all","ha-sync-mode":"automatic"}'

Setting policy "ha-all" for pattern "" to "{"ha-mode":"all","ha-sync-mode":"automatic"}" with priority "0" for vhost "/" ...

for vhost1 vhost
rabbitmqctl set_policy ha-all "" '{"ha-mode":"all","ha-sync-mode":"automatic"}' -p vhost1

Setting policy "ha-all" for pattern "" to "{"ha-mode":"all","ha-sync-mode":"automatic"}" with priority "0" for vhost "vhost1" ...