awk 截取某段时间的日志
好久没有截取nginx/haproxy 中 的日志了,竟有点不熟悉了。
记录一下,以免以后忘记。
NGINX 日志格式:
192.168.1.26 - - [14/Sep/2017:16:48:42 +0800] "GET /ui/favicons/favicon-16x16.png HTTP/1.1" 304 0 "http://192.168.1.124/app/kibana" "Mozilla/5.0 (Windows NT 10.0; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/59.0.3071.115 Safari/537.36" "-"
截取方法:
cat access.log | awk '$4 >="[14/Sep/2017:16:30:50" && $4 <="[14/Sep/2017:16:35:50"' >>/tmp/nginx-access.com
HAPROXY 日志格式:
Sep 28 10:17:45 localhost haproxy[12443]: 192.168.1.36:52935 [28/Sep/2017:10:17:45.950] Dianda server_new/192.168.0.19 0/0/0/1/1 304 237 - - ---- 604/96/0/0/0 0/0 "GET /images/logo.png HTTP/1.1"
截取方法:
cat haproxy.log |egrep "Sep 28"| awk '$3 >="10:08:00" && $3 <="10:18:00"' >/tmp/haproxy.txt
如果有些要过滤的可以考虑用 egrep
比如碰到这种格式
[2017-09-12 00:00:33.866] [INFO] access - 192.168.1.10 get api_server-121000-1505145633854 /v2/promotion/announcement 200 12ms
[2017-09-12 00:00:18.720] [INFO] access - 192.168.0.100 579 get api_server-579-1504195218720 /v3/user/coupon/list {} {} null
cat access.log-2017-08-31.log |awk '{print $9 , $NF}' |egrep -v 'null|\{|\}|N'|awk -F'ms' '{print $1}' |awk '$NF >= 1500'
执行结果, 访问哪个api 用时多少
/v2/order/query/type/count 2644
/v2/order/query/type/count 1901
/v2/order/query/type/count 2837
/v1/message/client/device 3673
/v2/order/place/new 1640
/v2/order/place 1890
/v1/message/client/device 1635
/v1/message/client/device 1667
/v1/message/client/device 1747
在段落前面加一行自定义数据
假设: a="11111111111111"
echo $a |awk '{for(i=1;i<=NF;i++){printf "--exclude="$i" "}{print ""}}'
结果变成:
--exclude=11111111111111