SAP DevOps笔试题

1.[Unix administration]
Please use Linux bash to fulfill the following Operation tasks.

• List all Java process
• List Linux kernel version
• Compress all log files under /var/log/nginx, and copy to server 10.0.0.1. User
  name:work, Password: Password

2.[Programming]
Please use script language(Python, Perl, Nodejs etc) to fulfill the following tasks in two ways

• Download a file from http://some.website/some.file
• Logon to https://some.website/login using POST http method, User name:work,Password: Password.

3.[Network]

• List >3 http method other than GET
• How to exam the DNS resolution
  e.g. www.baidu.com in Linux shell, List the possible commands you
  have know.

4.对B类网段172.168.0.0进行划分,子网掩码为255.255.192.0

• 要求计算可划分子网数
• 每个子网中主机数
• 确定每个子网的子网地址
• 确定每个子网合法IP地址
• 确定每个子网广播地址

---

参考答案

1.[Unix administration]
Please use Linux bash to fulfill the following Operation tasks.

• List all Java process

ps -ef | grep java

• List Linux kernel version

uname -r

• Compress all log files under /var/log/nginx, and copy to server 10.0.0.1. User
  name:work, Password: Password

分两步走：
首先, 执行查找所有的log文件，并压缩成为自命名的文件，假设为logzips.gz：
find /var/log/nginx -type f -name '*.log' -print | xargs tar cf - |gzip -c > ./logzips.gz 
然后，用scp进行copy，中途要输入被传输机器的密码，假设放到/var/mylog目录下：
scp ./logzips.gz work@10.0.0.1:/var/mylog/logzips.gz

2.[Programming]
Please use script language(Python, Perl, Nodejs etc) to fulfill the following tasks in two ways

• Download a file from http://some.website/some.file

[Python]方法1, 通过Http使用urllib模块：
import urllib 
print "downloading with urllib" 
url = 'http://some.website/some.file'  
urllib.urlretrieve(url, "some.file")

[Python]方法2，通过Http使用urllib2模块：
import urllib2 
print "downloading with urllib2" 
url = 'http://some.website/some.file' 
fin = urllib2.urlopen(url)
data = fin.read() 
with open("some.file", "wb") as fout:
    fout.write(data)

[Python]方法3， 通过使用requests模块
import requests 
print "downloading with requests"
url = 'http://some.website/some.file' 
response = requests.get(url) 
with open("some.file", "wb") as fout:
     fout.write(response.content)

• Logon to https://some.website/login using POST http method, User name:work,Password: Password.

方法1：不使用Cookie, 发送HTTP POST
import urllib2, urllib
data = {'Username' : 'work', 'Password' : 'Password'}
f = urllib2.urlopen(
        url     = 'https://some.website/login',
        data    = urllib.urlencode(data)
  )
print f.read()

方法2：使用CookieJar
import urllib2
import urllib
from cookielib import CookieJar

cj = CookieJar()
opener = urllib2.build_opener(urllib2.HTTPCookieProcessor(cj))
# input-type values from the html form
data = { 'Username' : 'work', 'Password' : 'Password' }
data_encoded = urllib.urlencode(data)
response = opener.open("https://some.website/login", data_encoded)
content = response.read()

3.[Network]

• List >3 http method other than GET

1. GET 从指定的url上获取内容
2. POST 提交body中的内容给服务器中指定的url中，属于非幂等的(non-idempotent)请求
3. HEAD 从指定的url上获取header内容(类似Get方式)
4. PUT 将body上传至服务器指定url处
5. DELETE 在指定url处删除资源
6. OPTIONS 获取指定url中能接收的请求方法
7. CONNECT 连接指定频段。当客户端需要通过代理服务器连接HTTPS服务器是用到。

具体参见HTTP Method

• How to exam the DNS resolution
  e.g. www.baidu.com in Linux shell, List the possible commands you
  have know.

原理都是依赖于一个很重的的文件/etc/resolv.conf
1. dig www.baidu.com (+tcp, +trace)
2. nslookup www.baidu.com
3. host www.baidu.com (用于查看某主机对应的IP）
4. ping www.baidu.com (如果ping不通，当然DNS解析就不对）

4.对B类网段172.168.0.0进行划分,子网掩码为255.255.192.0

• 要求计算可划分子网数

因为是B类网段，B类地址用16位表示网络ID，而其子网掩码是18位，则子网位数为18-16=2位，那么子网就有2^2=4个。

• 每个子网中主机数

IPv4总共用32位表示，分为网络ID和主机ID，其中子网掩码占了18位，即网络ID可用18位，则还剩32-18=14位来用于表示主机，即每个子网中主机数为2^14 -2 个 （由于特殊用途，丢弃全为0，和全为1这两种情况）。

• 确定每个子网的子网地址

子网地址则为每个网段的起始地址为全0即：
172.168.0.0
172.168.64.0
172.168.128.0
172.168.192.0

• 确定每个子网合法IP地址

172.168.0.1～172.168.63.254
172.168.64.1～172.168.127.254
172.168.128.1～172.168.191.254
172.168.192.1～172.168.255.254

• 确定每个子网广播地址

广播地址为每个网段的结束地址为全1即：
172.168.63.255
172.168.127.255
172.168.191.255
172.168.255.255

没有看明白的话，可以看看我的另一篇文章如何计算子网数