这几天又复习起安全评估了,重新操刀在一家运营商为几个业务系统进行安全评估工作,今天漏洞扫描时,记起一个小工具,共享一下。因为以往在进行漏洞扫描时,很偶然的情况下(虽说很偶然,但几个久经阵仗的老手都栽过跟头,我当然也不例外,这里面的血泪史,说起来,都是经验啊……😢),可能会有些脆弱的应用导致系统负载过高而当机。
监控的办法有很多,电信、移动或金融行业许多都有自身的监控平台,一旦业务异常便会有告警,这是一种形式。但有时无法实时查看告警平台时,就只能用土办法了,比如,持续地ping被扫描的服务器。
fping在这里能够起到很好的作用,我主要看中它的几个功能:
1、能够从文件导入一批IP地址或直接设置一个网段,同时ping;
2、回应timeout的时候,可以有声音提示;
3、可以设定间隔多长时间ping一次;
它的命令参数有:
Usage:
fping <host(-list)> [-s data_size] [-S size1/size2] [-c] [-t time] [-w timeout]
[-n count] [-h TTL] [-v TOS] [-r routes] [-R min/max] [-a] [-f]
[-b(-)] [-i] [-l] [-T] [-D] [-d ping_data] [-g host1/host2]
[-H filename] [-L filename]
Options:
-s : data_size in bytes up to 65500
-S : size sweep. Ping with size1, size1 + 1, ..., size 2 datalength
-c : continuous ping (higher priority than -n)
to see statistics and continue - type Control-Break;
to stop - type Control-C.
-t : time between 2 pings in ms up to 1000000
-w : timeout in ms to wait for each reply
-n : number of echo requests to send
-h : number of hops (TTL: 1 to 128)
-v : Type Of Service (0 to 255)
-r : record route (1 to 9 routes)
-R : random length between min and max (disabled when using -S)
-a : resolve addresses to hostnames
-f : set Don't Fragment flag in packet
-b : beep on every successful reply (- to beep on timeout)
-i : use ICMP dll instead of raw socket (disables -r)
-l : limit the output to ping results and errors
-T : print timestamp with each reply
-D : print datestamp and timestamp with each reply
-d : ping with specified data (disabled when using -R)
-g : ping IP range from host1 to host2
-H : get hosts from filename (comma delimited, filename with full path)
-L : logging to a text file
在评估报告中的“风险规避”这部份,把这张图贴进去,好歹也是种负责任的态度吧 ;)