#!/usr/bin/env python
# -*- coding: utf-8 -*-
"""
解析dmidecode命令输出结果，返回JSON格式数据
测试服务器Dell VMware虚拟机
测试系统为CentOS 6.x 7.x
Python版本为Python3.6
原版参考的是 https://pypi.org/project/dmidecode/ 但是在Python3上无法直接使用而且我觉得它有些逻辑问题需要做一些修改
"""
import sys
import subprocess
__version__ = "0.1.0"
__author__ = "rex.chen"
class Dmidecode(object):
def __init__(self):
# 注释掉里面的键值意味着忽略某些类型
self._DMI_TYPE = {
0: "BIOS Information", # BIOS信息 提供商、版本等
1: "System Information", # 系统信息包括服务器厂商、服务器的星号、服务器序列号
2: "Base Board Information",
3: "Chassis Information", # 可以获取服务器的高度，比如1U等。
4: "Processor Information", # 每个逻辑CPU的信息，物理CPU数量*核心数量=逻辑CPU数量
6: "Memory Module Information",
7: "Cache Information",
8: "Port Connector Information",
9: "System Slot Information",
10: "On Board Device Information",
11: "OEM Strings",
15: "System Event Log",
16: "Physical Memory Array",
17: "Memory Device", # 每个内存槽位上查的内存条信息，类型、容量、主频、序列号等
18: "32-bit Memory Error Information",
19: "Memory Array Mapped Address",
20: "Memory Device Mapped Address",
23: "System Reset",
24: "Hardware Security",
30: "Out-of-band Remote Access",
32: "System Boot Information",
33: "64-bit Memory Error Information",
38: "IPMI Device Information",
41: "Onboard Device",
# 126: "Inactive",
# 127: "End Of Table"
}
def parse_dmi(self, content):
"""
解析dmidecode命令输出
:param content: 传递进来dmidecode命令执行的原始输出结果
:return: 重新组装后的数据字典
"""
info = {}
try:
"""
这里是一个关键点，dmidecode命令输出其实是层级结构的，它使用制表符 来表示层级，lines可以列表，但后续处理会比较麻烦
所以这里使用迭代器一个是占用空间少，另外是你每一次传递lines到其他地方当调用next()或者for循环时它都是从上一次的位置
继续向下，这样对于处理这种dmidecode输出的有层级关系的非结构化数据比较方便。
"""
lines = iter(content.strip().splitlines())
while True:
try:
line = next(lines)
if line.startswith(b'Handle 0x'):
typ = int(str(line).split(',', 2)[1].strip()[len('DMI type'):])
if typ in self._DMI_TYPE:
if typ not in info:
info[self._DMI_TYPE[typ]] = []
info[self._DMI_TYPE[typ]].append(self._parse_handle_section(lines))
else:
info[self._DMI_TYPE[typ]].append(self._parse_handle_section(lines))
except StopIteration:
break
return info
except Exception as err:
print("Error occured in Function parse_dmi")
print(err)
def _parse_handle_section(self, lines):
"""
解析每个type下面的信息，也就是每一个 Handle 0x 下面的信息
:param lines: 传递所有的内容进来，也就是之前生成的迭代器，而且这个迭代器是接着上面的位置继续向后迭代
:return: 字典，每个type下面的子信息组成的字典
"""
data = {}
title = str(next(lines).strip().decode(encoding='utf-8'))
try:
for line in lines:
line = line.rstrip()
strline = str(line.strip().decode(encoding='utf-8'))
if line.startswith(b' '):
data[k].append(strline)
elif line.startswith(b' '):
k, v = str(line.strip().decode(encoding='utf-8')).split(":", 1)
if v:
data[k] = v.strip()
else:
data[k] = []
else:
break
except Exception as err:
print("Error occured in Function parse_handle_section")
print("Data section is %s " % title)
print(err)
return data
def main():
cmd = "sshpass -p 123456 ssh root@172.16.48.171 '/usr/sbin/dmidecode'"
mydmi = Dmidecode()
try:
completed_process = subprocess.run(cmd, shell=True, stdout=subprocess.PIPE, stderr=subprocess.PIPE)
if completed_process.returncode == 0:
data = completed_process.stdout
print(mydmi.parse_dmi(data))
else:
print("Returncode is not 0")
except Exception as err:
print(err)
if __name__ == "__main__":
try:
main()
finally:
sys.exit()
