监控软件之open-falcon

一、open-falcon介绍

1）中文社区介绍

http://book.open-falcon.org/zh_0_2/intro/

参照文档： https://www.cnblogs.com/LAlexH/p/11161943.html     

参照文档：https://www.cnblogs.com/straycats/p/7199209.html

视频连接：http://www.jikexueyuan.com/course/1651_3.html?ss=1

告警设置：https://www.cnblogs.com/python-lbl/p/10450186.html

2）falcon的优点

强大灵活的数据采集：自动发现，支持falcon-agent、snmp、支持用户主动push、用户自定义插件支持、opentsdb data model like（timestamp、endpoint、metric、key-value tags）
水平扩展能力：支持每个周期上亿次的数据采集、告警判定、历史数据存储和查询
高效率的告警策略管理：高效的portal、支持策略模板、模板继承和覆盖、多种告警方式、支持callback调用
人性化的告警设置：最大告警次数、告警级别、告警恢复通知、告警暂停、不同时段不同阈值、支持维护周期
高效率的graph组件：单机支撑200万metric的上报、归档、存储（周期为1分钟）
高效的历史数据query组件：采用rrdtool的数据归档策略，秒级返回上百个metric一年的历史数据
dashboard：多维度的数据展示，用户自定义Screen
高可用：整个系统无核心单点，易运维，易部署，可水平扩展
开发语言： 整个系统的后端，全部golang编写，portal和dashboard使用python编写

3）falcon的特性

数据采集方式多样灵活：支持agent、snmp、用户主动push、自定义插件等多种方式进行数据采集
高效率报警策略管理
人性化的告警设置
dashboard多维度数据展示
模板支持继承的同时支持覆盖策略项
server端无需做配置，只需要在client端按照agent则可以自动监控
引入tag概念，通过tag多维度对数据进行查询展示

4) falcon的架构图

Open-Falcon是一个比较大的分布式系统，有十几个组件。按照功能，这十几个组件可以划分为 基础组件、作图链路组件和报警链路组件，其安装部署的架构如下图所示

 二、open-falcon单机环境安装

1）安装redis

1.1）yum安装方式

wget -O /etc/yum.repos.d/CentOS-Base.repo http://mirrors.aliyun.com/repo/Centos-7.repo

或 yum install epel-release

yum install redis -y
systemctl start redis
systemctl enable redis
systemctl status redis

1.2）下载 tar包进行安装

#创建redis工作目录
mkdir /home/redis && cd /home/redis
#下载redis程序包：
wget http://download.redis.io/releases/redis-4.0.9.tar.gz
#编译安装
tar -zxvf redis-4.0.9.tar.gz
mv redis-4.0.9 redis4.0.9 && cd redis4.0.9
mkdir logs
make && make install
#修改配置文件
vim redis.conf
bind 0.0.0.0
daemonize yes
pidfile /var/run/redis_6379.pid
logfile "/home/redis/redis4.0.9/logs/redis.log"
#启动redis
redis-server /home/redis/redis4.0.9/redis.conf
#连接测试
redis-cli -h 127.0.0.1 -p 6379

 2）安装mysql

wget -i -c http://dev.mysql.com/get/mysql57-community-release-el7-10.noarch.rpm
yum -y install mysql57-community-release-el7-10.noarch.rpm
yum -y install mysql-community-server
systemctl start mysqld.service
[root@node01 ~]# grep "password" /var/log/mysqld.log
2019-07-13T02:39:54.602191Z 1 [Note] A temporary password is generated for root@localhost: i?5XuEqh+aRL
第一次登陆。必须给密码
mysql -uroot -pi?5XuEqh+aRL

mysql>  set global validate_password_policy=0;
Query OK, 0 rows affected (0.00 sec)

mysql> set global validate_password_length=1;
Query OK, 0 rows affected (0.00 sec)

mysql> ALTER USER 'root'@'localhost' IDENTIFIED BY '123456';
Query OK, 0 rows affected (0.00 sec)

systemctl stop mysqld.service
systemctl start mysqld.service
systemctl status mysqld.service
systemctl enable mysqld.service
mysql -uroot -p123456

 2.1) 卸载mysql源。防止每次yum操作都会自动更新

yum -y remove mysql57-community-release-el7-10.noarch

2.2）不要使用root账号

GRANT ALL  ON *.* TO 'falcon'@'localhost' IDENTIFIED BY 'falcon';
GRANT ALL  ON *.* TO 'falcon'@'%' IDENTIFIED BY 'falcon';
flush privileges;

2.1) 授权普通用户远程连接

授权远程连接
GRANT ALL PRIVILEGES ON *.* TO 'falcon'@'%' IDENTIFIED BY 'falconpassword' WITH GRANT OPTION;
flush privileges;

查看用户： SELECT DISTINCT CONCAT('User: ''',user,'''@''',host,''';') AS query FROM mysql.user;

部分版本授权不能使用 %，代表所有，可考虑使用 * 试一下

3）初始化表结构

git clone  https://github.com/open-falcon/falcon-plus.git
#导入表结构
cd ./falcon-plus/scripts/mysql/db_schema/
mysql -ufalcon -pfalcon < 1_uic-db-schema.sql
mysql -ufalcon -pfalcon < 2_portal-db-schema.sql
mysql -ufalcon -pfalcon < 3_dashboard-db-schema.sql
mysql -ufalcon -pfalcon < 4_graph-db-schema.sql
mysql -ufalcon -pfalcon < 5_alarms-db-schema.sql
#删除目录
rm -rf falcon-plus/

4）安装golang

#下载go安装包
https://dl.google.com/go/go1.12.7.linux-amd64.tar.gz
#解压至/home目录下
tar -zxvf go1.12.7.linux-amd64.tar.gz -C /home
#声明PATH
echo "export PATH=$PATH:/home/go/bin" >> /etc/profile
source  /etc/profile
#查看go版本
go version

 4.1）创建工作目录

export FALCON_HOME=/home
export WORKSPACE=$FALCON_HOME/open-falcon
mkdir -p $WORKSPACE

 5）在工作目录中下载安装包

#下载安装包
wget https://github.com/open-falcon/falcon-plus/releases/download/v0.2.0/open-falcon-v0.2.0.tar.gz
#解压
cd /home/open-falcon
tar  -zxvf open-falcon-v0.2.0.tar.gz

6）后端启动

#修改配置文件为自己设置的mysql用户和密码
grep -Ilr 3306  ./ | xargs -n1 -- sed -i 's/root:/falcon:falcon/g'
#启动服务
/home/open-falcon/open-falcon start
/home/open-falcon/open-falcon check
#显示如下则全部启动成功
　　falcon-graph UP 27685 
　　falcon-hbs UP 27697 
　　falcon-judge UP 27707 
　　falcon-transfer UP 27716 
　　falcon-nodata UP 27724 
　　falcon-aggregator UP 27732 
　　falcon-agent UP 27743 
　　falcon-gateway UP 27753 
　　falcon-api UP 27761 
　　falcon-alarm UP 28201

递归替换

grep -Ilr 3306  ./ | xargs -n1 -- sed -i 's/root:/falcon:falconpassword/g'
grep -Ilr 3306  ./ | xargs -n1 -- sed -i 's/127.0.0.1/172.20.16.5/g'

三、安装前端展示界面

1）下载展示模板 dashboard

#下载dashboard项目至本地
cd $WORKSPACE
git clone https://github.com/open-falcon/dashboard.git
#安装所需依赖包
yum install -y python-virtualenv
yum install -y python-devel
yum install -y openldap-devel
yum install -y mysql-devel
yum groupinstall "Development tools"

2）创建依赖环境

#创建独立的虚拟环境
cd $WORKSPACE/dashboard/
virtualenv ./env
#pip安装依赖
./env/bin/pip install -r pip_requirements.txt -i https://pypi.douban.com/simple

3）在依赖环境中创建配置文件

vim rrd/config
# TODO: read from api instead of db
PORTAL_DB_HOST = os.environ.get("PORTAL_DB_HOST","127.0.0.1")
PORTAL_DB_PORT = int(os.environ.get("PORTAL_DB_PORT",3306))
PORTAL_DB_USER = os.environ.get("PORTAL_DB_USER","falcon")
PORTAL_DB_PASS = os.environ.get("PORTAL_DB_PASS","falcon")
PORTAL_DB_NAME = os.environ.get("PORTAL_DB_NAME","falcon_portal")

# alarm database
# TODO: read from api instead of db
ALARM_DB_HOST = os.environ.get("ALARM_DB_HOST","127.0.0.1")
ALARM_DB_PORT = int(os.environ.get("ALARM_DB_PORT",3306))
ALARM_DB_USER = os.environ.get("ALARM_DB_USER","falcon")
ALARM_DB_PASS = os.environ.get("ALARM_DB_PASS","falcon")
ALARM_DB_NAME = os.environ.get("ALARM_DB_NAME","alarms")

4）启动服务查看状态

#启动
bash control start
bash control status
#查看日志
bash control tail

服务器安装完成

4.1）如果出现内部错误

[root@node01 dashboard]#  cat rrd/config.py   请查看改该文件连接的mysql是否正常加载了用户名和密码

四、客户端的安装

 1）从服务端拷贝文件到客户端

[root@node01 open-falcon]# pwd
/home/open-falcon
[root@node01 open-falcon]# scp -r agent/ root@192.168.1.7:/home/open-falcon/
[root@node01 open-falcon]# scp open-falcon root@192.168.1.7:/home/open-falcon/

 2）编辑配置文件

#编辑agent配置文件，修改hostname、transfer、heaetbeat配置项
vim agent/config/cfg.json
#启动agent，查看agent状态
./open-falcon start agent
./open-falcon check agent
tailf agent/logs/agent.log
#重载配置文件
curl 127.0.0.1:1988/config/reload

稍等片刻。机器自动发现

 五、查询基本使用

1）机器选择，监控指标选择

1.1）查看图像

2) Screen的功能的基本使用

归纳： 先创建demo组，再创建 相关监控的类。最后添加监控指标

 

再继续添加内存

 3）分组功能

添加机器

4）创建模板

添加监控策略

 将之前的主机组绑定模板

5）测试肯定会触发的报警值。模板里面进行修改

六、客户端的开机自启动

[root@iotansible0001 init.d]# pwd
/etc/rc.d/init.d
[root@iotansible0001 init.d]# cat falcon-agentd 
#!/bin/bash
# /etc/init.d/falcon-agentd
# chkconfig: 2345 20 80
# description: Starts and Stops falcon-agent

dir=/home/envuser/falcon
pid=`ps -ef | grep falcon-agent | grep -v falcon-agentd | grep -v "grep" | awk '{print $2}'`

case "$1" in
start)
        if [[ $pid -gt 0 ]];then
                echo $pid
                kill -9 $pid
                echo "Stopping falcon-agent ..."
        fi
        sleep 1
        echo "Starting falcon-agent ..."
        su - envuser -c "cd $dir && nohup ./open-falcon start agent &"
        ;;
stop)
        if [[ $pid -gt 0 ]];then
                echo $pid
                kill -9 $pid
                echo "Stopping falcon-agent ..."
                sleep 1
        else
                echo "Falcon-agent is stoped ..."
        fi
        ;;
restart)
        echo "Resstarting falcon-agent ..."
        if [[ $pid -gt 0 ]];then
                echo $pid        
                kill -9 $pid
                echo "Stopping falcon-agent ..."
        fi
        sleep 1
        echo "Starting falcon-agent ..."
        su - envuser -c "cd $dir && nohup ./open-falcon start agent &"
        ;;
*) 
        echo "Usage: falcon-agentd {start|stop|restart}" 
        exit 0 
esac 
exit 0

添加至启动项

chmod +x falcon-agentd
chkconfig --add falcon-agentd
chkconfig  falcon-agentd on

 七、客户端命令验证

[envuser@nginx-mqtt0001 bin]$ ./falcon-agent --check
net.if   ... ok
cpustat  ... ok
disk.io  ... ok
memory   ... ok
ss -s    ... ok
ss -tln  ... ok
kernel   ... ok
df.bytes ... ok
loadavg  ... ok
netstat  ... ok
ps aux   ... ok
du -bs   ... ok

八、推送监控数据

curl -X POST -d "[{"metric": "test_by_test", "endpoint": "test_by_test_ep", "timestamp": `date +%s`,"step": 60,"value": 1,"counterType": "GAUGE","tags": "region=test"}]" http://127.0.0.1:1988/v1/push &> /dev/null

九、 es 集群监控

引用配置

[es]
data_host = elk0001:9200,elk0002:9200,elk0003:9200
log_host = elk-log0001.eniot.io:9200,elk-log0002.eniot.io:9200,elk-log0003.eniot.io:9200

监控脚本

# coding: utf-8
import time
import datetime
import json
import traceback

from monitor_logger import Logger
from monitor_falcon import Falcon

from elasticsearch import Elasticsearch

log_file = u"eniot_monitor_es_status.log"

class ESstatus():
    def __init__(self,logger = None):
        self.logger = logger if logger else Logger(log_file).get_logger()
        self.falcon = Falcon(self.logger)

    def get_conf(self,cf):
        try:
            data_info = dict()
            region = cf.get(u"region", u"region")
            if not region:
                msg =u"get region by conf error!"
                self.logger.error(msg)
                return
            data_info.update({u"region": region})

            data_host = cf.get(u"es", u"data_host")
            if not data_host:
                msg =u"get es host data by conf error!"
                self.logger.error(msg)
                return
            data_info.update({u"data_host": data_host})

            log_host = cf.get(u"es", u"log_host")
            if not data_host:
                msg = u"get es host data by conf error!"
                self.logger.error(msg)
                return
            data_info.update({u"log_host": log_host})

            return data_info
        except:
            self.logger.error(traceback.format_exc())

    def push_falcon(self,region, excutetime ,  status,clusterName):
        try:
            endpoint = "eniot_monitor_es_status"

            metric = "eniot_monitor_es_status_excutetime"
            tags = "region={region},clusterName={clusterName}".format(
                region = region,
                clusterName = clusterName,
            )
            print(tags)
            falcon_push_data = self.falcon.get_push_data(endpoint, metric, tags, float(excutetime))
            self.falcon.push_data(falcon_push_data)

            metric = "eniot_monitor_es_status"

            falcon_push_data = self.falcon.get_push_data(endpoint, metric, tags, status)
            self.falcon.push_data(falcon_push_data)
        except:
            self.logger.error(traceback.format_exc())

    def monitor_es_client(self,region,host):
        try:
            esclient = Elasticsearch(host)
            start_time = time.clock()
            result = esclient.cat.health().split(" ")
            result_v = esclient.cat.health(v=True)
            print(result_v)
            clusterName =result[2]
            if result[3] != "green":
                status = 0
            else:
                status = 1

            end_time = time.clock()
            excutetime = end_time - start_time
            print("result = " + result[3])
            print("excutetime = " + str(excutetime))

            self.push_falcon(region, excutetime, status,clusterName)
        except:
            self.logger.error(traceback.format_exc())

    def main(self):
        try:
            cf =self.falcon.check_conf()
            data_info = self.get_conf(cf)
            if not data_info:
                msg = u"get es info error!"
                self.logger.warn(msg)
                return
            region = data_info["region"]
            log_host = data_info["log_host"].split(",")
            data_host = data_info["data_host"].split(",")
            self.monitor_es_client(region,log_host)
            self.monitor_es_client(region,data_host)

        except:
            self.logger.error(traceback.format_exc())

if __name__ == '__main__':
    app = ESstatus()
    app.main()