在现代企业级应用与大数据环境中，常用中间件如 Zookeeper、Kafka、Redis、Elasticsearch、Flink、Hadoop 以及 Doris 等，扮演着关键的角色，保证系统的高可用性、可扩展性和稳定运行。而 systemd 作为主流的 Linux 初始化系统和服务管理器，提供了统一、可靠的方式来管理这些中间件的启动、停止与自动重启。本文详细汇总了一系列常用中间件的 systemd 单元配置文件示例，覆盖直接运行进程和基于 Docker Compose 的容器化服务，帮助运维人员快速配置与管理各类中间件服务。通过规范的 systemd 配置，能够有效简化运维流程，提高服务启动效率，确保业务连续性。
zookeeper
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
| sudo tee /etc/systemd/system/zookeeper.service <<'EOF'
[Unit]
Description=zookeeper
After=network.target
[Service]
Type=forking
User=lbs
Group=lbs
KillMode=control-group
Environment="JAVA_HOME=/home/lbs/software/jdk8"
ExecStart=/home/lbs/software/zookeeper/bin/zkServer.sh start
ExecStop=/home/lbs/software/zookeeper/bin/zkServer.sh stop
SuccessExitStatus=0 143
PrivateTmp=false
LimitNOFILE=1000000
LimitNPROC=100000
TimeoutStopSec=10s
Restart=on-failure
RestartSec=30
[Install]
WantedBy=multi-user.target
EOF
|
kafka
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
| sudo tee /etc/systemd/system/kafka.service <<'EOF'
[Unit]
Description=kafka
After=zookeeper.service
Requires=zookeeper.service
[Service]
Type=forking
User=lbs
Group=lbs
KillMode=control-group
Environment="JAVA_HOME=/home/lbs/software/jdk8"
ExecStart=/home/lbs/software/kafka/bin/kafka-server-start.sh -daemon /home/lbs/software/kafka/config/server.properties
ExecStop=/home/lbs/software/kafka/bin/kafka-server-stop.sh
SuccessExitStatus=0 143
PrivateTmp=false
LimitNOFILE=1000000
LimitNPROC=100000
TimeoutStopSec=10s
Restart=on-failure
RestartSec=30
[Install]
WantedBy=multi-user.target
EOF
|
redis
节点1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
| sudo tee /etc/systemd/system/redis-6379.service <<'EOF'
[Unit]
Description=redis-6379
After=network.target
[Service]
Type=forking
User=lbs
Group=lbs
KillMode=control-group
ExecStart=/home/lbs/software/redis/bin/redis-server /home/lbs/software/redis/6379/conf/redis-6379.conf
ExecStop=/home/lbs/software/redis/bin/redis-cli -p 6379 -a Rongshu@2024 shutdown
SuccessExitStatus=0 143
PrivateTmp=false
LimitNOFILE=1000000
LimitNPROC=100000
TimeoutStopSec=10s
Restart=on-failure
RestartSec=30
[Install]
WantedBy=multi-user.target
EOF
|
节点2
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
| sudo tee /etc/systemd/system/redis-6380.service <<'EOF'
[Unit]
Description=redis-6380
After=network.target
[Service]
Type=forking
User=lbs
Group=lbs
KillMode=control-group
ExecStart=/home/lbs/software/redis/bin/redis-server /home/lbs/software/redis/6380/conf/redis-6380.conf
ExecStop=/home/lbs/software/redis/bin/redis-cli -p 6380 -a Rongshu@2024 shutdown
SuccessExitStatus=0 143
PrivateTmp=false
LimitNOFILE=1000000
LimitNPROC=100000
TimeoutStopSec=10s
Restart=on-failure
RestartSec=30
[Install]
WantedBy=multi-user.target
EOF
|
elasticSearch（es)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
| sudo tee /etc/systemd/system/es.service <<'EOF'
[Unit]
Description=elasticsearch
After=network.target
[Service]
Type=forking
User=lbs
Group=lbs
ExecStart=/home/lbs/software/elasticsearch/bin/elasticsearch -d
PrivateTmp=true
# 进程可以打开的最大文件数
LimitNOFILE=65535
# 进程可以打开的最大进程数
LimitNPROC=65535
# 最大虚拟内存
LimitAS=infinity
# 最大文件大小
LimitFSIZE=infinity
# 超时设置 0-永不超时
TimeoutStopSec=0
# SIGTERM是停止java进程的信号
KillSignal=SIGTERM
# 信号只发送给给JVM
KillMode=process
# java进程不会被杀掉
SendSIGKILL=no
# 正常退出状态
SuccessExitStatus=143
# 开机自启动
[Install]
WantedBy=multi-user.target
EOF
|
flink集群
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
| sudo tee /etc/systemd/system/flink-cluster.service <<'EOF'
[Unit]
Description=flink集群
After=network.target
[Service]
Type=forking
User=lbs
Group=lbs
KillMode=control-group
Environment="JAVA_HOME=/home/lbs/software/jdk8"
ExecStart=/home/lbs/software/flink/bin/start-cluster.sh
ExecStop=/home/lbs/software/flink/bin/stop-cluster.sh
SuccessExitStatus=0 143
PrivateTmp=false
LimitNOFILE=1000000
LimitNPROC=100000
TimeoutStopSec=10s
Restart=on-failure
RestartSec=30
[Install]
WantedBy=multi-user.target
EOF
|
kibana
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
| sudo tee /etc/systemd/system/kibana.service <<'EOF'
[Unit]
Description=Kibana Service
After=es.service
Requires=es.service
[Service]
Type=simple
User=lbs
Group=lbs
ExecStart=/home/lbs/software/kibana/bin/kibana
ExecStop=/usr/bin/kill -15 $MAINPID
ExecReload=/usr/bin/kill -HUP $MAINPID
RemainAfterExit=yes
PrivateTmp=true
# file size
LimitFSIZE=infinity
# cpu time
LimitCPU=infinity
# virtual memory size
LimitAS=infinity
# open files
LimitNOFILE=65535
# processes/threads
LimitNPROC=64000
# locked memory
LimitMEMLOCK=infinity
# total threads (user+kernel)
TasksMax=infinity
TasksAccounting=false
[Install]
WantedBy=multi-user.target
EOF
|
hadoop集群
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
| sudo tee /etc/systemd/system/hadoop-cluster.service <<'EOF'
[Unit]
Description=hadoop-cluster
After=network.target
[Service]
Type=forking
User=lbs
Group=lbs
KillMode=control-group
Environment="JAVA_HOME=/home/lbs/software/jdk8"
ExecStart=/home/lbs/software/hadoop/sbin/start-all.sh
ExecStop=/home/lbs/software/hadoop/sbin/stop-all.sh
SuccessExitStatus=0 143
PrivateTmp=false
LimitNOFILE=1000000
LimitNPROC=100000
TimeoutStopSec=10s
Restart=on-failure
RestartSec=30
[Install]
WantedBy=multi-user.target
EOF
|
doris
fe
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
| sudo tee /etc/systemd/system/doris-fe.service <<'EOF'
[Unit]
Description=doris-fe
After=network.target
[Service]
Type=forking
User=lbs
Group=lbs
KillMode=control-group
Environment="JAVA_HOME=/home/lbs/software/jdk8"
ExecStart=/home/lbs/software/doris/fe/bin/start_fe.sh --daemon
ExecStop=/home/lbs/software/doris/fe/bin/stop_fe.sh --daemon
SuccessExitStatus=0 143
PrivateTmp=false
LimitNOFILE=1000000
LimitNPROC=100000
TimeoutStopSec=10s
Restart=on-failure
RestartSec=30
[Install]
WantedBy=multi-user.target
EOF
|
be
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
| sudo tee /etc/systemd/system/doris-be.service <<'EOF'
[Unit]
Description=doris-be
After=network.target
[Service]
Type=forking
User=lbs
Group=lbs
KillMode=control-group
Environment="JAVA_HOME=/home/lbs/software/jdk8"
ExecStart=/home/lbs/software/doris/be/bin/start_be.sh --daemon
ExecStop=/home/lbs/software/doris/be/bin/stop_be.sh --daemon
SuccessExitStatus=0 143
PrivateTmp=false
LimitNOFILE=1000000
LimitNPROC=100000
TimeoutStopSec=10s
Restart=on-failure
RestartSec=30
[Install]
WantedBy=multi-user.target
EOF
|
broker
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
| sudo tee /etc/systemd/system/doris-broker.service <<'EOF'
[Unit]
Description=doris-broker
After=network.target
[Service]
Type=forking
User=lbs
Group=lbs
KillMode=control-group
Environment="JAVA_HOME=/home/lbs/software/jdk8"
ExecStart=/home/lbs/software/doris/extensions/apache_hdfs_broker/bin/start_broker.sh --daemon
ExecStop=/home/lbs/software/doris/extensions/apache_hdfs_broker/bin/stop_broker.sh --daemon
SuccessExitStatus=0 143
PrivateTmp=false
LimitNOFILE=1000000
LimitNPROC=100000
TimeoutStopSec=10s
Restart=on-failure
RestartSec=30
[Install]
WantedBy=multi-user.target
EOF
|
nacos
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
| sudo tee /etc/systemd/system/nacos.service <<'EOF'
[Unit]
Description=nacos
After=network.target
[Service]
Type=forking
User=lbs
Group=lbs
KillMode=control-group
Environment="JAVA_HOME=/home/lbs/software/jdk8"
ExecStart=/home/lbs/software/nacos/bin/startup.sh
ExecStop=/home/lbs/software/nacos/bin/shutdown.sh
SuccessExitStatus=0 143
PrivateTmp=false
LimitNOFILE=1000000
LimitNPROC=100000
TimeoutStopSec=10s
Restart=on-failure
RestartSec=30
[Install]
WantedBy=multi-user.target
EOF
|
harbor
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
| sudo tee /etc/systemd/system/harbor.service <<'EOF'
[Unit]
Description=harbor
After=docker.service
Requires=docker.service
[Service]
Type=simple
Restart=on-failure
RestartSec=5
ExecStart=/usr/local/bin/docker-compose -f /root/software/harbor/harbor/docker-compose.yml up
ExecStop=/usr/local/bin/docker-compose -f /root/software/harbor/harbor/docker-compose.yml down
[Install]
WantedBy=multi-user.target
EOF
|
mysql
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
| sudo tee /etc/systemd/system/mysql.service <<'EOF'
[Unit]
Description=mysql
After=docker.service
Requires=docker.service
[Service]
Type=simple
Restart=on-failure
RestartSec=5
ExecStart=/usr/local/bin/docker-compose -f /opt/1panel/apps/mysql/mysql/docker-compose.yml up
ExecStop=/usr/local/bin/docker-compose -f /opt/1panel/apps/mysql/mysql/docker-compose.yml down
[Install]
WantedBy=multi-user.target
EOF
|
portainer
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
| sudo tee /etc/systemd/system/portainer.service <<'EOF'
[Unit]
Description=portainer
After=docker.service
Requires=docker.service
[Service]
Type=simple
Restart=on-failure
RestartSec=5
ExecStart=/usr/local/bin/docker-compose -f /home/lbs/software/portainer/docker-compose.yaml up
ExecStop=/usr/local/bin/docker-compose -f /home/lbs/software/portainer/docker-compose.yaml down
[Install]
WantedBy=multi-user.target
EOF
|
prometheus等
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
| sudo tee /etc/systemd/system/prometheus.service <<'EOF'
[Unit]
Description=prometheus, grafana, alertmanager, prometheus-alert, pushgateway
After=docker.service
Requires=docker.service
[Service]
Type=simple
Restart=on-failure
RestartSec=5
ExecStart=/usr/local/bin/docker-compose -f /home/lbs/software/prometheus/docker-compose.yaml up
ExecStop=/usr/local/bin/docker-compose -f /home/lbs/software/prometheus/docker-compose.yaml down
[Install]
WantedBy=multi-user.target
EOF
|
jenkins
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
| sudo tee /etc/systemd/system/jenkins.service <<'EOF'
[Unit]
Description=jenkins
After=docker.service
Requires=docker.service
[Service]
Type=simple
Restart=on-failure
RestartSec=5
ExecStart=/usr/local/bin/docker-compose -f /home/lbs/software/jenkins/jenkins/docker-compose.yaml up
ExecStop=/usr/local/bin/docker-compose -f /home/lbs/software/jenkins/jenkins/docker-compose.yaml down
[Install]
WantedBy=multi-user.target
EOF
|
sonarqube
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
| sudo tee /etc/systemd/system/sonarqube.service <<'EOF'
[Unit]
Description=sonarqube, postgres
After=docker.service
Requires=docker.service
[Service]
Type=simple
Restart=on-failure
RestartSec=5
ExecStart=/usr/local/bin/docker-compose -f /home/lbs/software/sonarqube/sonarqube/docker-compose.yaml up
ExecStop=/usr/local/bin/docker-compose -f /home/lbs/software/sonarqube/sonarqube/docker-compose.yaml down
[Install]
WantedBy=multi-user.target
EOF
|
node_exporter
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
| sudo tee /etc/systemd/system/node_exporter.service <<'EOF'
[Unit]
Description=node_exporter
After=network.target
[Service]
User=lbs
Group=lbs
ExecStart=/home/lbs/software/exporter/node_exporter/node_exporter
Restart=on-failure
RestartSec=20
[Install]
WantedBy=multi-user.target
EOF
|
flume
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
| sudo tee /etc/systemd/system/flume.service <<'EOF'
[Unit]
Description=flume
After=network.target
[Service]
Type=forking
User=lbs
Group=lbs
KillMode=control-group
Environment="JAVA_HOME=/home/lbs/software/jdk8"
ExecStart=/home/lbs/software/flume/bin/flume-test.sh start
ExecStop=/home/lbs/software/flume/bin/flume-test.sh stop
SuccessExitStatus=0 143
PrivateTmp=false
LimitNOFILE=1000000
LimitNPROC=100000
TimeoutStopSec=10s
Restart=on-failure
RestartSec=30
[Install]
WantedBy=multi-user.target
EOF
|
结语
本文介绍的常用中间件 systemd 管理配置文件，覆盖了从传统服务到容器化部署的多种场景，适用于多种中间件系统的大规模集群环境。合理配置 systemd 服务单元，不仅能够实现服务的自动化管理和故障恢复，还能通过配置限制和环境变量提升系统安全性与性能表现。建议在实际使用中，根据具体实例路径和业务需求，灵活调整配置内容，并结合日志管理与监控手段，保障中间件服务的平稳运行。希望本文对运维同学在日常管理和自动化运维方面有所助益，提升整体运维效率和系统稳定性。
