Question

linux shell 想写个脚本，让tcpdump命令抓包并写入文件，执行指定时间后结束进程，然后再开始新一轮抓包

脚本如下：
#!/bin/bash
datetime=`date +%F-%H:%M:%S`
i=1
while [ $i -lt 5 ]
do
echo "start NO.$i capture data"
tcpdump -i eth2 -w ./$datetime.cap &
tdid=`pgrep tcpdump`
sleep 7s
echo "$tdid"
kill -9 $tdid
let i++
done
结果只是生产了一个抓包数据文件“2015-07-09-20:21:25.cap”，执行结果如下：
start NO.1 capture data
tcpdump: WARNING: eth2: no IPv4 address assigned
tcpdump: listening on eth2, link-type EN10MB (Ethernet), capture size 96 bytes
2828
start NO.2 capture data
tcpdump: WARNING: eth2: no IPv4 address assigned
tcpdump: listening on eth2, link-type EN10MB (Ethernet), capture size 96 bytes
./test.sh: line 14: 2828 Killed tcpdump -i eth2 -w ./$datetime.cap
2828
5687
./test.sh: line 12: kill: (2828) - No such process
start NO.3 capture data
tcpdump: WARNING: eth2: no IPv4 address assigned
tcpdump: listening on eth2, link-type EN10MB (Ethernet), capture size 96 bytes
./test.sh: line 14: 5687 Killed tcpdump -i eth2 -w ./$datetime.cap
10111
start NO.4 capture data
tcpdump: WARNING: eth2: no IPv4 address assigned
tcpdump: listening on eth2, link-type EN10MB (Ethernet), capture size 96 bytes
./test.sh: line 14: 10111 Killed tcpdump -i eth2 -w ./$datetime.cap
10111
12486
./test.sh: line 12: kill: (10111) - No such process
请问哪里有问题？或是怎么写这个需求的脚本？

Answer 1

datetime=`date +%F-%H:%M:%S` 这首源个者山态去掉

抓包改成唯纯
tcpdump -i eth2 -w ./`date +%F-%H:%M:%S` &

或者把 datetime=`date +%F-%H:%M:%S` 放到循环里面

追问

试了下，是可以循环生产文件了，但只有结束脚本后，最后一个文件有数据，除最后 一个文件其他的文件所是空的，这是为什么？

追答

一边ping着，一边执行脚本