2. Environment Preparation#

Component	Version
OS	Debian GNU/Linux 12 (bookworm)

Smokeping Deployment#

Smokeping is an open-source tool for monitoring network status and stability (developed by the author of rrdtool). Smokeping sends various types of test packets to target devices and systems, measures and records the results, and visualizes network latency using rrdtool-based charts. This makes it easy to understand the real-time state of network communication, such as latency, packet loss rate, and whether the connection is multi-line BGP.

Components Overview#

general (general settings), alerts (alert settings), Database (database parameters), presentation (custom network views), slaves (Smokeping slaves), targets (target configuration — including probe/plugin definitions)

Docker Deployment#

1
docker pull linuxserver/smokeping
2
docker run -d \
3
--name=smokeping \
4
-e TZ=Asia/Shanghai \
5
-p 9080:80 \
6
-v /home/config/smokeping:/config \
7
-v /home/data/smokeping:/data \
8
--restart unless-stopped \
9
linuxserver/smokeping

Deploy the Push Script#

1
apt update && apt upgrade
2
apt install -y rrdtool
3
cd /home/config/smokeping/
4
git clone https://github.com/Sayyiku/idc_ping_monitor.git
5
unzip idc_ping_monitor-master.zip(如果是因为网络上传，需要用到此命令解压)
6
mv idc_ping_monitor-master/smokeping/location/ /home/config/smokeping/

Edit Smokeping Monitoring Configuration#

1
vim Targets

1
*** Targets ***
2

3
probe = FPing
4

5
menu = Top
6
title = 网络节点质量监控
7
remark = Smokeping 网络质量监控系统
8

9
@include /config/location/cmcc
10
@include /config/location/telcom
11
@include /config/location/tencent
12
@include /config/location/unicom

Refresh Smokeping#

1
find /home/data/smokeping/ -name "*.rrd"|xargs rm -f
2
docker restart 容器 ID

Modify the Script#

Change IP to the local host IP.

1
root@VM-4-17-debian:/home/config/smokeping# vim /home/config/smokeping/idc_ping_monitor-master/collection_to_prometheus.py
2

3
'LOG_FILE' : '/tmp/smoking_pushgateway.log',
4
'prometheus_gateway' : 'http://ip:9091',
5
'data_dir' : '/home/data/smokeping'

Configure a Cron Job#

1
crontab -e
2
* * * * * python3 /home/config/smokeping/idc_ping_monitor-master/collection_to_prometheus.py

Prometheus Deployment#

Introduction to Time Series Data#

A time series database focuses on time as the primary axis and records how things change as time passes. A relational database mainly describes relationships between entities, such as one-to-many or many-to-many, and these relationships are relatively complex compared to time. Time is simpler: you only need to record what changed for a given entity at a specific point in time. Monitoring data fits this pattern very well: we simply record how a metric changes as time goes by.

Time series databases are particularly suitable for handling “change”. In a relational database, relationships are represented using table schemas. Once a schema is defined, data is continuously collected according to that schema. If something changes—such as adding a new metric dimension—you must modify the schema. DDL operations are costly for relational databases. Time series databases do not have this problem: time series datasets track system changes by continuously inserting new data instead of updating existing records. Therefore, for monitoring data, collecting new metric dimensions does not affect any overall “schema”—there effectively is no rigid table schema to modify.

Deploy Prometheus in a Container#

1
root@VM-4-17-debian:~# docker run -d -p 9090:9090 --name prometheus prom/prometheus
2
root@VM-4-17-debian:~# docker cp prometheus:/etc/prometheus/prometheus.yml ./

Persistent Deployment#

1
root@VM-4-17-debian:~# docker rm -f prometheus
2
prometheus
3
root@VM-4-17-debian:~# mkdir /data/prometheus/
4
root@VM-4-17-debian:~# cp prometheus.yml /data/prometheus/
5
root@VM-4-17-debian:~# docker run -d -p 9090:9090 --name prometheus \
6
-v /data/prometheus/:/etc/prometheus/ \
7
-e TZ="Asia/Shanghai" \
8
-v /etc/localtime:/etc/localtime \
9
prom/prometheus
10
3fbee8038e5256c622dda73078bd5e04aea3c7abe9a2da62b61cf64541d0c50e
11
root@VM-4-17-debian:~#

Prometheus Pushgateway Deployment#

Docker Deployment#

1
root@VM-4-17-debian:~# docker run -d --name="prometheus_pushgateway" -p 9091:9091 prom/pushgateway

Modify Prometheus Configuration to Add Pushgateway#

1
root@VM-4-17-debian:~# vim /data/prometheus/prometheus.yml
2
  - job_name: prometheus_pushgateway
3
    honor_labels: true                       # 避免收集数据本身的 job 和 instance被pushgateway实例信息覆盖
4
    static_configs:
5
    - targets: ['ip:9091']
6
      labels:
7
        instance: pushgateway
8
root@VM-4-17-debian:~# docker restart prometheus

Prometheus Web (IP:9090)#

Check whether the Pushgateway status is UP.

Grafana Deployment#

Official docs: https://grafana.com/docs/grafana/latest/

Docker Deployment#

1
[root@VM-4-17-debian ~]# mkdir /data/grafana-storage
2
[root@VM-4-17-debian ~]# chmod 777 /data/grafana-storage/
3

4
[root@VM-4-17-debian ~]# docker run -d -p 3000:3000 \
5
--name grafana \
6
-v /data/grafana-storage:/var/lib/grafana \
7
-e "GF_SECURITY_ADMIN_PASSWORD=123456" \
8
grafana/grafana