nats Cluster deployment

explain ： The cluster is deployed on three machines , Catalogue of each machine 、 Installation package 、 The configuration name is the same , But configure the of each node ip Different from routing , Pay attention to the corresponding changes , If deployed on one machine , You need to change the port of each node , Avoid port conflicts

1. download

NATS Server The latest version is v2.8.4 edition , There are three binary distribution packages , See ：https://github.com/nats-io/nats-server/releases

• nats-server-v2.8.4-darwin-amd64.tar.gz 　for Mac Version of
• nats-server-v2.8.4-linux-amd64.tar.gz 　　for Linux x64 Version of
• nats-server-v2.8.4-linux-arm.tar.gz for Linux ARM Version of

2. decompression

mkdir /biz-code/nats/ -p
cd /biz-code/nats/
wget https://github.com/nats-io/nats-server/releases/download/v2.8.4/nats-server-v2.8.4-linux-amd64.tar.gz
tar xf nats-server-v2.8.4-linux-amd64.tar.gz
cd nats-server-v2.8.4-linux-amd64/

3. Add the configuration

vim gnastd.conf

Node one ：

port: 4333
http: 6333
net: 0.0.0.0

cluster {
    
    host: 172.16.0.15
    port: 8333
# authorization {
    
# user: root
# password: 123456
# timeout: 1
# }

    routes = [
        nats-route://172.16.0.18:8333
        nats-route://172.16.0.24:8333
    ]
}
debug:   true
trace:   true
log_file: "/biz-code/nats/gnatsd-4222.log"
# max_connections
max_connections: 100
# Duration the server can block on a socket write to a client. Exceeding the
# deadline will designate a client as a slow consumer.
#  Test for problems 
write_deadline: "2s"

Node two ：

port: 4333
http: 6333
net: 0.0.0.0

cluster {
    
    host: 172.16.0.18
    port: 8333
# authorization {
    
# user: root
# password: 123456
# timeout: 1
# }

    routes = [
        nats-route://172.16.0.15:8333
        nats-route://172.16.0.24:8333
    ]
}
debug:   true
trace:   true
log_file: "/biz-code/nats/gnatsd-4222.log"
# max_connections
max_connections: 100
# Duration the server can block on a socket write to a client. Exceeding the
# deadline will designate a client as a slow consumer.
#  Test for problems 
write_deadline: "2s"

Node three ：

port: 4333
http: 6333
net: 0.0.0.0

cluster {
    
    host: 172.16.0.24
    port: 8333
# authorization {
    
# user: root
# password: 123456
# timeout: 1
# }

    routes = [
        nats-route://172.16.0.18:8333
        nats-route://172.16.0.15:8333
    ]
}
debug:   true
trace:   true
log_file: "/biz-code/nats/gnatsd-4222.log"
# max_connections
max_connections: 100
# Duration the server can block on a socket write to a client. Exceeding the
# deadline will designate a client as a slow consumer.
#  Test for problems 
write_deadline: "2s"

There is a problem ： The account and password for cluster communication authentication are written in the configuration file , However, in the process of debugging, the error has been reported and the authentication has failed , Then I annotated it , The account and password specified by parameters at startup

4. start-up

./nats-server --config gnastd.conf --user root --pass 123456

5. Check the log

cd /biz-code/nats
tail -f gnatsd-4222.log

If it doesn't appear in the log ERR, It is estimated that the startup is successful , In addition, after debugging, you can set the debug: true Option comment