Zabbix integration Cloud Alert (Rui like cloud) to achieve telephone messages warning

Zabbix integration Duo platform as the Smart Alarm CA (Cloud Alert)

I. Introduction and preliminary understanding

Cloud Alert through the application, alarm access control system / platform, centralized management of your alarm, unified dispatch notification, unified analysis.

The platform is the first to understand and use in the second half of 2017, before the name is called oneitsM. Warning Product name: OneAlertNow the product has been migrated to the cloud as the Core, and renamed CloudAlert. This article is the use of the product was introduced to early warning and record.

We must first register an account: official website link , then visit our account. Our choice Cloud Alert.

Then click on the top of our integration.

We can directly integrate the industry's leading monitoring tools, such as: Zabbix, Nagios, Prometheus, OpenFalcon, SolarWinds etc., but also can be integrated via Email mailbox or the REST API access to your alarms.

Second, integrated into Zabbix

We need to get our first APPkey, and then in the installation, the incoming parameters.

First, install Agent

Switch to the zabbixscripts directory (see how zabbixthe scripts directory):

cd /usr/local/zabbix-server/share/zabbix/alertscripts

Gets Cloud Alert Agent package:

wget https://download.aiops.com/ca_agent/zabbix/ca_zabbix_release-2.1.0.tar.gz

Decompression, installation.

tar -xzf ca_zabbix_release-2.1.0.tar.gz 
cd cloudalert/bin 
bash install.sh APPkey

Note: 1, depending on the installation during the installation prompts, enter zabbix management address, administrator user name and password.

2, zabbix management address is correct example: http://zabbix.server.com/zabbix

It represents a successful installation when prompted "successful installation"!
In the zabbix servermanagement interface to see if the added successfully.

1561622519970

The script appears it means success.

Third, create and assign policies to inform policy

这里要严重的注意一点,就是 CloudAlert 的预警级别只有三种,它和 zabbix 的级别对应见下表.我们在设置通知策略和分派策略需要注意.

zabbix 级别状态	参数值	OneAlert 级别状态
information (信息)	1	提醒
not_classified (未分类)	1	提醒
warning (警告)	2	警告
average (一般严重)	2	警告
high (严重)	3	严重
disaster (灾难)	3	严重

分派策略

通知策略

我们这里只使用到 CloudAlert 的通知策略的通知方式中的电话和短信,因为我们微信已经对接了我们的企业微信预警,邮箱也使用了企业邮箱(进垃圾邮箱的概率更低一些).所以我们这里只设置通知方式为电话和短信. 并且只有在严重预警的时候才会触发.

四、设置动作

在执行安装脚本的时候,默认已经帮我们添加好了动作 cloudalert action。但是没有设置触发条件,我们可以设置一下触发条件,触发器示警度大于等于严重的时候,进行触发该动作.

五、生成预警

当我们的预警达到阈值,就会触发报警.

短信预警内容:

【睿象云】16:48:53,发生严重级别告警:Zabbix agent on Test-186 is unreachable for 5 minutesTest-186 Agent ping:Up (1)Zabbix agent on Test-186 is unreachable for 5 minutes,告警对象:Test-186,告警编号:585679

六、卸载Cloud Alert

Web 设置卸载

删除报警媒介 cloudalert media。
删除用户群组 cloudalert group。
删除用户 cloudalert。
删除动作 cloudalert action。

有人会说我们没有添加上面的东西，为什么会存在上面这些内容，我们在执行 install.sh 脚本的时候就会添加这些.

脚本文件卸载

删除脚本删除 /usr/local/zabbix-server/share/zabbix/alertscripts 的cloudalert文件夹。

七、注意事项

错误内容

我安装的 zabbix server 是使用的 docker 安装的, 脚本目录是使用的数据卷(我们可以用 docker inspect container_id 查看到)。我们将脚本放置在该数据卷后，安装，也提示成功了，但是在预警的时候，有报错。

报错内容如下：

/usr/lib/zabbix/alertscripts/cloudalert/bin
/usr/lib/zabbix/alertscripts/cloudalert/bin/log.sh: line 11: /var/lib/docker/volumes/bb221b74a7d8ad528194867830db0c1ac8fdc31f2ab0ee4456ffce61646fd83a/_data/cloudalert/logs/cloudalert.log: No such file or directory
cp: cannot stat '/var/lib/docker/volumes/bb221b74a7d8ad528194867830db0c1ac8fdc31f2ab0ee4456ffce61646fd83a/_data/cloudalert/logs/cloudalert.log': No such file or directory
/usr/lib/zabbix/alertscripts/cloudalert/bin/log.sh: line 16: /var/lib/docker/volumes/bb221b74a7d8ad528194867830db0c1ac8fdc31f2ab0ee4456ffce61646fd83a/_data/cloudalert/logs/cloudalert.log: No such file or directory
% Total % Received % Xferd Average Speed Time Time Time Current
Dload Upload Total Spent Left Speed

0 0 0 0 0 0 0 0 --:--:-- --:--:-- --:--:-- 0
100 623 0 89 100 534 1141 6850 --:--:-- --:--:-- --:--:-- 6935
/usr/lib/zabbix/alertscripts/cloudalert/bin/log.sh: line 11: /var/lib/docker/volumes/bb221b74a7d8ad528194867830db0c1ac8fdc31f2ab0ee4456ffce61646fd83a/_data/cloudalert/logs/cloudalert.log: No such file or directory
cp: cannot stat '/var/lib/docker/volumes/bb221b74a7d8ad528194867830db0c1ac8fdc31f2ab0ee4456ffce61646fd83a/_data/cloudalert/logs/cloudalert.log': No such file or directory
/usr/lib/zabbix/alertscripts/cloudalert/bin/log.sh: line 16: /var/lib/docker/volumes/bb221b74a7d8ad528194867830db0c1ac8fdc31f2ab0ee4456ffce61646fd83a/_data/cloudalert/logs/cloudalert.log: No such file or directory

解析过程与解决问题

从上面我们可以看到是脚本 log.sh 执行的过程中报错了，提示没有该文件，它写入的文件是 /var/lib/docker/volumes/bb221b74a7d8ad528194867830db0c1ac8fdc31f2ab0ee4456ffce61646fd83a/_data/cloudalert/logs/cloudalert.log ，这个文件路径是 docker 宿主机的日志文件路径，程序在 docker 里面运行，这个路径肯定是获取不到的。我们通过查看 log.sh 脚本发现,

#!/bin/bash
if [ -z "$DIR" ]; then
    DIR="$( cd "$( dirname "$0"  )" && pwd  )"
fi

source $DIR/cloudalert.conf
function log(){
  path=$base_path
  log=$path/logs/cloudalert.log
  time=`date +%Y-%m-%d\ %H:%M:%S`
  echo $time $1 [$2]: "$3" >> $log
  bak_log=$path/logs/cloudalert.log_`date +%Y-%m`
  if [ ! -f $bak_log ];
  then
    cp $log $bak_log
    > $log
  fi
}

日志路径引用了 base_path,这个值是在 cloudalert.conf 里定义的. 我们可以在配置文件中看到, base_path 值获取的是宿主机 cloudalert 所在的路径. 而不是实际在 docker 容器里的路径。

current_path=/var/lib/docker/volumes/bb221b74a7d8ad528194867830db0c1ac8fdc31f2ab0ee4456ffce61646fd83a/_data/cloudalert/bin
base_path=/var/lib/docker/volumes/bb221b74a7d8ad528194867830db0c1ac8fdc31f2ab0ee4456ffce61646fd83a/_data/cloudalert
agentVersion=1130
AppKey=1233444555666
zabbix_host=http://127.0.0.1
zabbix_url=http://127.0.0.1/api_jsonrpc.php
user=Admin
password=admin

我们将 base_path 的值更改为在docker 容器里面的值 /usr/lib/zabbix/alertscripts/cloudalert .

然后就没有报错了.

反思

这个问题的引起，是我在宿主机下将执行的该脚本，导致的脚本执行的时候获取的是宿主机的目录，而不是 docker 主机里面的目录路径，在 docker 容器里面执行脚本，即可避免该问题。