通过Python利用saltstack进行生成服务器资产清单

Posted 2020-11-03

(以下代码Linux测试成功)

linux-node0.oldboyedu.com  192.168.1.30 安装salt-master,salt-minion

linux-node1.oldboyedu.com  192.168.1.31 安装salt-minion

这里主要用到saltstack的grains,就是saltstack minion端生成的一些静态信息，比如CPU，内存，主机名什么的，而这些就是我们所需要的

执行salt \* grains.items,会打印一大堆的默认抓取的信息，其中一部分，如下

 

我们当然只挑我们需要的，操作如下

 

获取主机名

salt linux-node1.oldboyedu.com grains.item host

 

获取IP地址

salt linux-node1.oldboyedu.com grains.item ipv4

 

获取CPU核数

salt \* grains.item num_cpus

 

以此类推，根据自己所需，提取~~~

 

值得注意的是，grains信息里面并没有硬盘信息，所以还需通过disk.usage这个选项，得到我们所需的硬盘信息

执行salt linux-node1.oldboyedu.com disk.usage

                                        

其中1K-blocks即我们所需的硬盘信息，根据需求只需要盘/,所以后面就会计算这个盘的配额

 

11.py代码：

 

#coding=utf-8

import salt.client as sc

import json

 

###salt调用

local = sc.LocalClient()

###目标主机指定

tgt = "*"

 

###获取grains，disk信息

grains = local.cmd(tgt,"grains.items")

diskusage = local.cmd(tgt,"disk.usage")

 

###主要应用列表即文件开头

app_name = ["tomcat","zookeeper","redis","mysql","nginx"]

cols = "Hostname,IP address,RAM(GB),CPU num,Operation,DATA /(GB),project"

 

###打开一个.csv文件，以便写入

ret_file = open("ret.csv","w")

###首先写入开头，有点字段名的意思

ret_file.write(cols + "\n")

try:

    for i in grains.keys():

###打印信息可注释掉

        print grains[i]["nodename"]

        print "ipv4" + ":" ,grains[i]["ipv4"]

        print "mem_total" + ":" , grains[i]["mem_total"] / 1024 + 1

        print "num_cpus" + ":" , grains[i]["num_cpus"]

        print "osfullname" + ":" , grains[i]["osfullname"]

        print "release" + ":" , grains[i]["lsb_distrib_id"]

###万一主机没有/ 数据盘1048576是1024x102

        if "/" not in diskusage[i]:

            print "diskusage" + ":" + "have no /data disk"

        else:

            data_vol = int(diskusage[i]["/"]["1K-blocks"])

            print "diskusage" + ":" , data_vol / 1048576

###去掉127.0.0.1这个地址

        ipv4 = str(grains[i]["ipv4"]).replace(", '127.0.0.1'","")

        

###因为一些历史遗留问题，这里取得不是主机名，而是salt-minion的id名，用以判断主要应用

 

        hostname = grains[i]["id"]

        ipv4 = str(grains[i]["ipv4"]).replace(", '127.0.0.1'","")

        ipv4 = ipv4.replace(",","and")

        mem = grains[i]["mem_total"] / 1024 + 1

        num_cpu = grains[i]["num_cpus"]

        OS = grains[i]["osfullname"] + grains[i]["lsb_distrib_id"]

        if "/" not in diskusage[i]:

            disk_data = "None"

        else:

            disk_data = data_vol / 1048576

###通过minion ID名来判断主要运行服务，比如xx-mysql-1，则运行mysql

        for j in app_name:

            if j in hostname.lower():

                app =  j

                break

            else:

                app = "undefined"

        c = ","

        

 ###连接并写入

        line = hostname + c + ipv4 + c + str(mem) + c + str(num_cpu) + c + str(OS) + c + str(disk_data) + c + app

        ret_file.write(line + "\n")

except Exception,e:

    print "Exception:\n",e

finally:

    ret_file.close()

 

 

 执行salt \* grains.items,发现"lsb_distrib_id"代替了原来的"lsb_distrib_release"，所以上面用了"lsb_distrib_id"

 

 

[[email protected] ~]# python 11.py

linux-node0

ipv4: ['127.0.0.1', '192.168.1.30', '192.168.1.38']

mem_total: 1

num_cpus: 1

osfullname: CentOS Linux

release: CentOS Linux

diskusage: 16

linux-node1

ipv4: ['127.0.0.1', '192.168.1.31', '192.168.122.1']

mem_total: 3

num_cpus: 1

osfullname: CentOS Linux

release: CentOS Linux

diskusage: 16

 

 

 

 

 

 

 

 

执行后在同一目录下生成ret.csv文件，打开如下：