傳統部署方式
1、純手工scp
2、純手工登錄git pull 、svn update
3、純手工xftp往上拉
4、開發給打一個壓縮包，rz上去。解壓
傳統部署缺點：
1、全程運維參與，占用大量時間
2、上線速度慢。
3、認為失誤多。管理混亂
4、回滾慢，不及時
新項目上線，規划排在第一位
一般銀行都不提供測試接口。比如一些電商公司測試的話，經常把商品調節成1分，只能特定賬號能看到。
環境的規划
1、開發環境-開發者本地有自己的環境，然后運維需要設置的開發環境，放的是大家共用的服務。如開發數據庫mysql，其它：redis、Memcached。
2、測試環境：功能測試環境和性能測試環境
3、預生產環境：一般可以用生產環境中的某個節點擔任
4、生產環境：直接對用戶提供服務的環境
預生產環境產生的原因：
1、數據庫不一致：測試環境和生產環境數據庫肯定不一樣的。
2、使用生產環境的聯調接口。例如，支付接口
預生產環境--生產環境--灰度發布
灰度發布：
阿里雲產品上線，都是一個區一個區上的。肯定不是一下子都上的
qq彈窗：恭喜你獲得某某版本資格，請下載新版本享用，你就是小白鼠，這個也是一種灰度發布
規划
已經有一個可以上線的代碼在代碼倉庫。我們如何設計一套生產自動化部署系統。
1、規划
2、實現
3、總結和擴展。PDCA
4、在生產環境應用
自動化部署系統的規划
需求：
1個集群有10個節點。一鍵部署這10個節點。
2、一鍵回滾到任意版本
3、一鍵回滾到上個版本
部署：
1、代碼放在哪里：svn，git
2、獲取什么版本代碼？
svn+git直接拉去某個分支
svn:指定版本號
git:指定tag
3、差異解決：
（1）、各個節點直接差異：
（2）、代碼倉庫和實際的差異。配置文件是否在代碼倉庫中
（3）、配置文件未必一樣：crontab.xml預生產節點
4、如何更新。java tomcat。需要重啟。
5、測試。
6、串行和並行 分組部署
7如何執行。（1）shell執行。（2）web界面
關於配置文件存放：
配置文件放代碼倉庫里，豈不是所有開發都能連數據庫了。
因此配置文件不能放git里。
有的人把配置文件放某個分支里。讓一些人沒法看到
我覺得可以單獨有個放配置文件的git
騰訊藍鯨:我幫你做個平台，你寫個腳本，我幫你發布到某個機器上，你通過平台幫你執行
一些公司現在的運維，不管任何發布，都是做好界面，讓項目負責人去通過界面管理
關於差異文件：
可能有些節點有特殊的差異性文件
自動化部署流程設計
自動化部署實戰-shell函數
環境准備
系統版本
|
1
2
3
4
5
|
[root@linux-node1 ~]
# cat /etc/redhat-release
CentOS Linux release 7.1.1503 (Core)
[root@linux-node1 ~]
# uname -rm
3.10.0-229.el7.x86_64 x86_64
[root@linux-node1 ~]
#
|
主機名和IP
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
|
node1
[root@linux-node1 ~]
# hostname
linux-node1.nmap.com
[root@linux-node1 ~]
# cat /etc/hosts
127.0.0.1 localhost localhost.localdomain localhost4 localhost4.localdomain4
::1 localhost localhost.localdomain localhost6 localhost6.localdomain6
192.168.56.11 linux-node1 linux-node1.nmap.com
192.168.56.12 linux-node2 linux-node2.nmap.com
[root@linux-node1 ~]
#
node2
[root@linux-node2 ~]
# hostname
linux-node2.nmap.com
[root@linux-node2 ~]
# cat /etc/hosts
127.0.0.1 localhost localhost.localdomain localhost4 localhost4.localdomain4
::1 localhost localhost.localdomain localhost6 localhost6.localdomain6
192.168.56.11 linux-node1 linux-node1.nmap.com
192.168.56.12 linux-node2 linux-node2.nmap.com
[root@linux-node2 ~]
#
|
兩台web服務器，node1和node2作為兩個web服務器，同時node1也作為部署分發服務器，去管理2個node節點上的web包
兩個節點添加普通用戶www，作為web服務器管理用戶。
|
1
2
3
4
5
6
7
8
9
|
[root@linux-node1 scripts]
# useradd -u 1001 www
[root@linux-node1 scripts]
# id www
uid=1001(www) gid=1001(www)
groups
=1001(www)
[root@linux-node1 scripts]
#
[root@linux-node2 ~]
# useradd -u 1001 www
[root@linux-node2 ~]
# id www
uid=1001(www) gid=1001(www)
groups
=1001(www)
[root@linux-node2 ~]
#
|
配置www用戶登錄其他機器不用密碼。密鑰認證。以后www用戶作為管理其它機器的用戶
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
|
[root@linux-node1 .
ssh
]
# su - www
[www@linux-node1 ~]$
ssh
-keygen -t rsa
Generating public
/private
rsa key pair.
Enter
file
in
which
to save the key (
/home/www/
.
ssh
/id_rsa
):
Created directory
'/home/www/.ssh'
.
Enter passphrase (empty
for
no passphrase):
Enter same passphrase again:
Your identification has been saved
in
/home/www/
.
ssh
/id_rsa
.
Your public key has been saved
in
/home/www/
.
ssh
/id_rsa
.pub.
The key fingerprint is:
70:37:ff:d0:17:e0:74:1d:c9:04:28:bb:de:ec:1f:7f www@linux-node1.nmap.com
The key's randomart image is:
+--[ RSA 2048]----+
| .++++|
| . .o oo.|
| . . = . . |
| o o o . .|
| S . o . .|
| . o . |
| . o .. |
| . o o E|
| .... ..|
+-----------------+
[www@linux-node1 ~]$
|
查看公鑰
|
1
2
3
4
5
6
7
8
9
10
11
|
[www@linux-node1 ~]$
cd
.
ssh
/
[www@linux-node1 .
ssh
]$ ll
total 8
-rw------- 1 www www 1679 Apr 5 03:41 id_rsa
-rw-r--r-- 1 www www 406 Apr 5 03:41 id_rsa.pub
[www@linux-node1 .
ssh
]$
cat
id_rsa.pub
ssh
-rsa
AAAAB3NzaC1yc2EAAAADAQABAAABAQDcZh8EEk2
/rS6B/tLHnRpZGrGIJYFHg7zRFvuT3N9jvOFhYJdWv
+8WSQuT0pvxNM4eR0N5Ma9wVvKPo
/lVjCaFK
+M0dENJVhi6m9OKMtoo2u
jvvuyinNPP4pyoK6ggG5jOlEkHoLcbWCRG
/j3pN1rZYV
+1twET9xi2IA4UQkgPvKKYWjq7NUR0v5BWsgEQt7VvjcLWTlltTVeGb3FDVKIjDnioIBmLmVwJS64N+GGgAj5YQ+bKHTwY
anEMD39JGKxo0RXTZB5sa734yfNjc3hTZXB4RCcGdzgcMJs
/Rt5VeZ277zF86xr4Hd5cioAbV6Y1RvELjmpvrqUUz3tcaKId
www@linux-node1.nmap.com
[www@linux-node1 .
ssh
]$
|
node2也添加node1的公鑰
改成600權限才能正常登錄
|
1
2
3
4
5
6
7
8
9
|
[www@linux-node2 ~]$
cd
.
ssh
/
[www@linux-node2 .
ssh
]$ vim authorized_keys
[www@linux-node2 .
ssh
]$
cat
authorized_keys
ssh
-rsa
AAAAB3NzaC1yc2EAAAADAQABAAABAQDcZh8EEk2
/rS6B/tLHnRpZGrGIJYFHg7zRFvuT3N9jvOFhYJdWv
+8WSQuT0pvxNM4eR0N5Ma9wVvKPo
/lVjCaFK
+M0dENJVhi6m9OKMtoo2u
jvvuyinNPP4pyoK6ggG5jOlEkHoLcbWCRG
/j3pN1rZYV
+1twET9xi2IA4UQkgPvKKYWjq7NUR0v5BWsgEQt7VvjcLWTlltTVeGb3FDVKIjDnioIBmLmVwJS64N+GGgAj5YQ+bKHTwY
anEMD39JGKxo0RXTZB5sa734yfNjc3hTZXB4RCcGdzgcMJs
/Rt5VeZ277zF86xr4Hd5cioAbV6Y1RvELjmpvrqUUz3tcaKId
www@linux-node1.nmap.com
[www@linux-node2 .
ssh
]$
chmod
600 authorized_keys
[www@linux-node2 .
ssh
]$
|
登錄測試--成功
|
1
2
3
|
[www@linux-node1 .
ssh
]$
ssh
192.168.58.12
Last login: Mon Apr 10 00:31:23 2017 from 192.168.58.11
[www@linux-node2 ~]$
|
讓node1的www用戶ssh自己也不需要輸入密碼。
node1添加公鑰
本地也放自己的密鑰，這樣可以假裝模擬成3台機器。2個ssh免密鑰的機器
node1同時作為部署機
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
|
[www@linux-node1 .
ssh
]$ ll
total 12
-rw------- 1 www www 1679 Apr 5 03:41 id_rsa
-rw-r--r-- 1 www www 406 Apr 5 03:41 id_rsa.pub
-rw-r--r-- 1 www www 175 Apr 5 03:43 known_hosts
[www@linux-node1 .
ssh
]$ vim authorized_keys
[www@linux-node1 .
ssh
]$
chmod
600 authorized_keys
[www@linux-node1 .
ssh
]$
ssh
192.168.58.11
The authenticity of host
'192.168.58.11 (192.168.58.11)'
can't be established.
ECDSA key fingerprint is 8b:4e:2f:
cd
:37:89:02:60:3c:99:9f:c6:7a:5a:29:14.
Are you sure you want to
continue
connecting (
yes
/no
)?
yes
Warning: Permanently added
'192.168.58.11'
(ECDSA) to the list of known hosts.
Last login: Wed Apr 5 03:40:47 2017
[www@linux-node1 ~]$
exit
logout
Connection to 192.168.58.11 closed.
[www@linux-node1 .
ssh
]$
ssh
192.168.58.11
Last login: Wed Apr 5 03:46:21 2017 from 192.168.58.11
[www@linux-node1 ~]$
|
開始寫自動化部署腳本
根據上面的流程圖，先把大體框架寫出來
先把框架寫出來，然后每個函數里寫echo
看看腳本執行流程是否有問題
code_diff 拷貝差異部署文件
這是面向過程的一種開發方式
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
|
[root@linux-node1 ~]
# mkdir /scripts -p
[root@linux-node1 ~]
# cd /scripts/
[root@linux-node1 scripts]
# vim deploy.sh
[root@linux-node1 scripts]
# chmod +x deploy.sh
[root@linux-node1 scripts]
# ./deploy.sh
Usage: .
/deploy
.sh [ deploy | rollback ]
[root@linux-node1 scripts]
# cat deploy.sh
#!/bin/bash
#Shell Env
SHELL_NAME=
"deploy.sh"
SHELL_DIR=
"/home/www"
SHELL_LOG=
"${SHELL_DIR}/${SHELL_NAME}.log"
#Code Env
CODE_DIR=
"/deploy/code/deploy"
CONFIG_DIR=
"/deploy/config"
TMP_DIR=
"/deploy/tmp"
TAR_DIR=
"/deploy/tar"
usage(){
echo
$
"Usage: $0 [ deploy | rollback ]"
}
code_get(){
echo
code_get
}
code_build(){
echo
code_build
}
code_config(){
echo
code_config
}
code_tar(){
echo
code_tar
}
code_scp(){
echo
code_scp
}
cluster_node_remove(){
echo
cluster_node_remove
}
code_deploy(){
echo
code_deploy
}
config_diff(){
echo
config_diff
}
code_test(){
echo
code_test
}
cluster_node_in(){
echo
cluster_node_in
}
rollback(){
echo
rollback
}
main(){
case
$1
in
deploy)
code_get;
code_build;
code_config;
code_tar;
code_scp;
cluster_node_remove;
code_deploy;
config_diff;
code_test;
cluster_node_in;
;;
rollback)
rollback;
;;
*)
usage;
esac
}
main $1
[root@linux-node1 scripts]
#
|
最末尾還要加個main，否則無法執行
腳本再優化下
盡量不要讓$1來回傳，否則可能會亂
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
|
main(){
DEPLOY_METHOD=$1
case
$DEPLOY_METHOD
in
deploy)
code_get;
code_build;
code_config;
code_tar;
code_scp;
cluster_node_remove;
code_deploy;
config_diff;
code_test;
cluster_node_in;
;;
|
繼續完善腳本--添加日志和鎖
1、凡是不記錄日志的腳本就是刷流氓，執行到哪一步失敗的啊？
2、腳本是否可以多個人一起執行？（最好不要多個人一起執行）不允許多人執行的話可以上鎖
一般鎖文件放下面目錄下
|
1
2
3
4
|
[root@linux-node1 ~]
# cd /var/run/lock/
[root@linux-node1 lock]
# ls
iscsi lockdev lvm ppp subsys
[root@linux-node1 lock]
#
|
我們可以單獨添加個目錄，給它用，因為權限問題，需要授權改變屬組，我們使用tmp目錄
主函數執行之前，應該先判斷鎖文件是否存在，執行的時候也應該生成這個lock文件
既然2個地方用到了它，是否可以把它制作成變量
新的腳本如下，主要添加了鎖的功能
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
|
[root@linux-node1 scripts]
# cat deploy.sh
#!/bin/bash
#Shell Env
SHELL_NAME=
"deploy.sh"
SHELL_DIR=
"/home/www"
SHELL_LOG=
"${SHELL_DIR}/${SHELL_NAME}.log"
#Code Env
CODE_DIR=
"/deploy/code/deploy"
CONFIG_DIR=
"/deploy/config"
TMP_DIR=
"/deploy/tmp"
TAR_DIR=
"/deploy/tar"
LOCK_FILE=
"/tmp/deploy.lock"
usage(){
echo
$
"Usage: $0 [ deploy | rollback ]"
}
shell_lock(){
touch
${LOCK_FILE}
}
shell_unlock(){
rm
-f ${LOCK_FILE}
}
code_get(){
echo
code_get
sleep
60;
}
code_build(){
echo
code_build
}
code_config(){
echo
code_config
}
code_tar(){
echo
code_tar
}
code_scp(){
echo
code_scp
}
cluster_node_remove(){
echo
cluster_node_remove
}
code_deploy(){
echo
code_deploy
}
config_diff(){
echo
config_diff
}
code_test(){
echo
code_test
}
cluster_node_in(){
echo
cluster_node_in
}
rollback(){
echo
rollback
}
main(){
if
[ -f ${LOCK_FILE} ];
then
echo
"Deploy is running"
&&
exit
;
fi
DEPLOY_METHOD=$1
case
$DEPLOY_METHOD
in
deploy)
shell_lock;
code_get;
code_build;
code_config;
code_tar;
code_scp;
cluster_node_remove;
code_deploy;
config_diff;
code_test;
cluster_node_in;
shell_unlock;
;;
rollback)
shell_lock;
rollback;
shell_unlock;
;;
*)
usage;
esac
}
main $1
[root@linux-node1 scripts]
#
|
先執行下檢查語法錯誤
|
1
2
3
4
5
6
7
8
9
10
11
12
13
|
[root@linux-node1 scripts]
# ./deploy.sh deploy
code_get
code_build
code_config
code_tar
code_scp
cluster_node_remove
code_deploy
config_diff
code_test
cluster_node_in
[root@linux-node1 scripts]
# ./deploy.sh rollback
rollback
|
加個sleep測試下鎖的功能
給一個函數加下sleep 測試下執行中，另外的人是否可以執行這個腳本
|
1
2
3
4
|
code_get(){
echo
code_get
sleep
60;
}
|
運行腳本
|
1
2
|
[root@linux-node1 scripts]
# ./deploy.sh deploy
code_get
|
新打開一個窗口執行測試
正常情況下一個窗口執行部署，再開一個窗口肯定執行不了
|
1
2
3
4
5
6
7
|
[root@linux-node1 scripts]
# ./deploy.sh deploy
Deploy is running
[root@linux-node1 scripts]
# ./deploy.sh deploy
Deploy is running
[root@linux-node1 scripts]
# ./deploy.sh rollback
Deploy is running
[root@linux-node1 scripts]
#
|
增加日志功能
其實就是echo一行到日志文件中，每個函數寫加echo 寫到日志里，這樣比較low
能不能寫個日志函數，加時間戳。以后日志函數可以復制到其它腳本里
|
1
2
3
4
5
|
[www@linux-node1 scripts]$
date
"+%Y-%m-%d"
2017-04-23
[www@linux-node1 scripts]$
date
"+%H-%M-%S"
22-10-34
[www@linux-node1 scripts]$
|
腳本默認從上到下執行，遇到函數先加載，但是不執行
繼續優化下。如下
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
|
#!/bin/bash
# Date/Time Veriables
CDATE=$(
date
"+%Y-%m-%d"
)
CTIME=$(
date
"+%H-%M-%S"
)
#Shell Env
SHELL_NAME=
"deploy.sh"
SHELL_DIR=
"/home/www"
SHELL_LOG=
"${SHELL_DIR}/${SHELL_NAME}.log"
#Code Env
CODE_DIR=
"/deploy/code/deploy"
CONFIG_DIR=
"/deploy/config"
TMP_DIR=
"/deploy/tmp"
TAR_DIR=
"/deploy/tar"
LOCK_FILE=
"/tmp/deploy.lock"
|
還不能這么寫，不然以后的時間都是一樣的
可以改成這樣，它不會執行
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
|
#!/bin/bash
# Date/Time Veriables
CDATE=
'date "+%Y-%m-%d"'
CTIME=
'date "+%H-%M-%S"'
#Shell Env
SHELL_NAME=
"deploy.sh"
SHELL_DIR=
"/home/www"
SHELL_LOG=
"${SHELL_DIR}/${SHELL_NAME}.log"
#Code Env
CODE_DIR=
"/deploy/code/deploy"
CONFIG_DIR=
"/deploy/config"
TMP_DIR=
"/deploy/tmp"
TAR_DIR=
"/deploy/tar"
LOCK_FILE=
"/tmp/deploy.lock"
|
打包的時候，也用到時間戳命名了。還得用一個固定不變的時間用於打包
因為解壓的時候，scp的時候用必須知道確定的包名字。
這里用到了2個時間，log-date是讓它不執行的，cdate是讓它執行的
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
|
#!/bin/bash
# Date/Time Veriables
LOG_DATE=
'date "+%Y-%m-%d"'
LOG_TIME=
'date "+%H-%M-%S"'
CDATE=$(
date
"+%Y-%m-%d"
)
CTIME=$(
date
"+%H-%M-%S"
)
#Shell Env
SHELL_NAME=
"deploy.sh"
SHELL_DIR=
"/home/www"
SHELL_LOG=
"${SHELL_DIR}/${SHELL_NAME}.log"
#Code Env
CODE_DIR=
"/deploy/code/deploy"
CONFIG_DIR=
"/deploy/config"
TMP_DIR=
"/deploy/tmp"
TAR_DIR=
"/deploy/tar"
|
自己先測試下
|
1
2
3
4
5
6
7
|
[root@linux-node1 ~]
# LOG_DATE='date "+%Y-%m-%d"'
[root@linux-node1 ~]
# LOG_TIME='date "+%H-%M-%S"'
[root@linux-node1 ~]
# echo $LOG_DATE
date
"+%Y-%m-%d"
[root@linux-node1 ~]
# echo $LOG_TIME
date
"+%H-%M-%S"
[root@linux-node1 ~]
#
|
eval的用法
|
1
2
3
4
5
|
[root@linux-node1 ~]
# eval $LOG_TIME
22-21-05
[root@linux-node1 ~]
# eval $LOG_DATE
2017-04-23
[root@linux-node1 ~]
#
|
怎么可以讓它在一行呢。暫時沒找到辦法（倒是可以單獨定義一個時間變量）
|
1
2
3
4
|
[root@linux-node1 ~]
# eval $LOG_DATE && eval $LOG_TIME
2017-04-23
22-22-48
[root@linux-node1 ~]
#
|
單獨定義一個時間變量（這里用不到，但是可以實現）
|
1
2
3
4
5
6
|
[root@linux-node1 ~]
# D_T='date "+%Y-%m-%d-%H-%M-%S"'
[root@linux-node1 ~]
# echo $D_T
date
"+%Y-%m-%d-%H-%M-%S"
[root@linux-node1 ~]
# eval $D_T
2017-04-26-19-33-01
[root@linux-node1 ~]
#
|
編寫記錄日志函數
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
|
usage(){
echo
$
"Usage: $0 [ deploy | rollback ]"
}
writelog(){
LOGINFO=$1
echo
"${CDATE} ${CTIME}: ${SHELL_NAME} : ${LOGINFO}"
>> ${SHELL_LOG}
}
shell_lock(){
touch
${LOCK_FILE}
}
shell_unlock(){
rm
-f ${LOCK_FILE}
}
|
這樣code_get函數就記錄日志了
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
|
shell_lock(){
touch
${LOCK_FILE}
}
shell_unlock(){
rm
-f ${LOCK_FILE}
}
code_get(){
writelog code_get;
}
code_build(){
echo
code_build
}
|
規范點，加上雙引號，分號可要可不要，寫上不會報錯
|
1
2
3
4
5
6
7
8
9
10
11
|
shell_unlock(){
rm
-f ${LOCK_FILE}
}
code_get(){
writelog
"code_get"
;
}
code_build(){
echo
code_build
}
|
獲取代碼
把代碼放哪里？
為什么創建這些目錄，寫着寫着你就知道了
|
1
2
3
4
5
6
7
8
9
10
11
12
|
[root@linux-node1 scripts]
# mkdir /deploy/config -p
[root@linux-node1 scripts]
# mkdir /deploy/tmp -p
[root@linux-node1 scripts]
# mkdir /deploy/tar -p
[root@linux-node1 scripts]
# mkdir /deploy/code -p
[root@linux-node1 scripts]
# cd /deploy/
[root@linux-node1 deploy]
# ll
total 0
drwxr-xr-x 2 root root 6 Apr 23 22:37 code
drwxr-xr-x 2 root root 6 Apr 23 22:37 config
drwxr-xr-x 2 root root 6 Apr 23 22:37
tar
drwxr-xr-x 2 root root 6 Apr 23 22:37 tmp
[root@linux-node1 deploy]
#
|
最終目錄建立成這種
|
1
2
3
4
5
6
7
8
9
10
11
12
13
|
[root@linux-node1 deploy]
# cd code/
[root@linux-node1 code]
# mkdir web-demo -p
[root@linux-node1 code]
# cd ..
[root@linux-node1 deploy]
# tree
.
├── code
│ └── web-demo
├── config
├──
tar
└── tmp
5 directories, 0 files
[root@linux-node1 deploy]
#
|
修改腳本
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
|
#Shell Env
SHELL_NAME=
"deploy.sh"
SHELL_DIR=
"/home/www"
SHELL_LOG=
"${SHELL_DIR}/${SHELL_NAME}.log"
#Code Env
CODE_DIR=
"/deploy/code/web-demo"
CONFIG_DIR=
"/deploy/config"
TMP_DIR=
"/deploy/tmp"
TAR_DIR=
"/deploy/tar"
LOCK_FILE=
"/tmp/deploy.lock"
usage(){
echo
$
"Usage: $0 [ deploy | rollback ]"
}
|
有的不需要編譯，拉下代碼之后，可以先把配置文件放進去
|
1
2
3
4
|
code_get(){
writelog
"code_get"
;
cd
$CODE_DIR && git pull
}
|
配置文件不要放這個目錄下，這個目錄只用來更新---git pull.你不好判斷配置文件是倉庫里面的，還是你專門下載下來的(最佳實踐)
規划的時候，只讓這里目錄執行git pull
下面這個目錄用於整合配置文件。
|
1
|
TMP_DIR=
"/deploy/tmp"
|
繼續優化獲取代碼的函數
|
1
2
3
4
5
|
code_get(){
writelog
"code_get"
;
cd
$CODE_DIR && git pull
cp
-r ${CODE_DIR} ${TMP_DIR}/
}
|
配置操作的函數時候，覺得不合適，應該區分項目，標准化。比如web-demo可以理解為一個項目包名字
|
1
2
3
4
5
6
|
#Code Env
CODE_DIR=
"/deploy/code/web-demo"
CONFIG_DIR=
"/deploy/config/web-demo"
TMP_DIR=
"/deploy/tmp"
TAR_DIR=
"/deploy/tar"
LOCK_FILE=
"/tmp/deploy.lock"
|
目錄新建
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
|
[root@linux-node1 scripts]
# cd /deploy/
[root@linux-node1 deploy]
# cd config/
[root@linux-node1 config]
# mkdir web-demo
[root@linux-node1 config]
# cd ..
[root@linux-node1 deploy]
# tree
.
├── code
│ └── web-demo
├── config
│ └── web-demo
├──
tar
└── tmp
6 directories, 0 files
[root@linux-node1 deploy]
#
|
模擬下，在里面寫上hehe
|
1
2
3
4
5
6
|
[root@linux-node1 deploy]
# cd config/
[root@linux-node1 config]
# cd web-demo/
[root@linux-node1 web-demo]
# vim config.ini
[root@linux-node1 web-demo]
# cat config.ini
hehe
[root@linux-node1 web-demo]
#
|
因為web-demo項目出現頻繁，把它弄成變量
|
1
2
3
4
5
6
7
|
#Code Env
PRO_NAME=
"web-demo"
CODE_DIR=
"/deploy/code/web-demo"
CONFIG_DIR=
"/deploy/config/web-demo"
TMP_DIR=
"/deploy/tmp"
TAR_DIR=
"/deploy/tar"
LOCK_FILE=
"/tmp/deploy.lock"
|
調整下腳本，優化code_config函數
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
|
code_get(){
writelog
"code_get"
;
cd
$CODE_DIR && git pull
cp
-r ${CODE_DIR} ${TMP_DIR}/
}
code_build(){
echo
code_build
}
code_config(){
echo
code_config
/bin/cp
-r $CONFIG_DIR/* $TMP_DIR/$PRO_NAME
}
|
規范下，給變量加大括號
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
|
code_get(){
writelog
"code_get"
;
cd
$CODE_DIR && git pull
cp
-r ${CODE_DIR} ${TMP_DIR}/
}
code_build(){
echo
code_build
}
code_config(){
echo
code_config
/bin/cp
-r ${CONFIG_DIR}/* ${TMP_DIR}/
"${PRO_NAME}"
}
|
注意是/bin/cp ，這樣原先有配置文件，這里可以直接替換了
如果開發把配置文件打包進去了。連接的是測試的庫，假如你部署生產環境了，連接測試的庫。出了問題，誰背黑鍋
運維是最后一道防線。開發和測試沒遇到。你背黑鍋
該給包重命名了
|
1
2
3
4
5
|
code_config(){
echo
code_config
/bin/cp
-r ${CONFIG_DIR}/* ${TMP_DIR}/
"${PRO_NAME}"
PKG_NAME=
"${PRO_NAME}"
_
"${API_VER}"
-
"${CDATE}-${CTIME}"
}
|
繼續優化
|
1
2
3
4
5
6
|
code_config(){
echo
code_config
/bin/cp
-r ${CONFIG_DIR}/* ${TMP_DIR}/
"${PRO_NAME}"
PKG_NAME=
"${PRO_NAME}"
_
"${API_VER}"
-
"${CDATE}-${CTIME}"
cd
${TMP_DIR} &&
mv
${PRO_NAME} ${PKG_NAME}
}
|
添加版本號，先隨便定義個版本
|
1
2
3
4
5
6
|
code_get(){
writelog
"code_get"
;
cd
$CODE_DIR && git pull
cp
-r ${CODE_DIR} ${TMP_DIR}/
API_VER=
"123"
}
|
現在沒有git pull 假裝以echo 代替git pull
|
1
2
3
4
5
6
|
code_get(){
writelog
"code_get"
;
cd
$CODE_DIR &&
echo
"git pull"
cp
-r ${CODE_DIR} ${TMP_DIR}/
API_VER=
"123"
}
|
屬組授權
|
1
2
|
[root@linux-node1 scripts]
# chown -R www:www /deploy/
[root@linux-node1 scripts]
#
|
內容為hehe,生成代碼（頁面）
|
1
2
3
4
5
|
[root@linux-node1 scripts]
# cd /deploy/code/web-demo/
[root@linux-node1 web-demo]
# echo hehe>>index.html
[root@linux-node1 web-demo]
# cat index.html
hehe
[root@linux-node1 web-demo]
#
|
文件和目錄結構如下
|
1
2
3
4
5
6
7
8
9
10
11
12
13
|
[root@linux-node1 deploy]
# tree
.
├── code
│ └── web-demo
│ └── index.html
├── config
│ └── web-demo
│ └── config.ini
├──
tar
└── tmp
6 directories, 2 files
[root@linux-node1 deploy]
#
|
以www用戶測試腳本
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
|
[root@linux-node1 deploy]
# cd /scripts/
[root@linux-node1 scripts]
# chown -R www:www /scripts/deploy.sh
[root@linux-node1 scripts]
# ll
total 12
-rw-r--r-- 1 root root 234 Apr 3 23:51 cobbler_list.py
-rw-r--r-- 1 root root 1533 Apr 4 00:01 cobbler_system_api.py
-rwxr-xr-x 1 www www 1929 Apr 23 23:04 deploy.sh
[root@linux-node1 scripts]
# su - www
Last login: Sun Apr 23 22:06:44 CST 2017 on pts
/0
[www@linux-node1 scripts]$ .
/deploy
.sh deploy
git pull
code_build
code_config
code_tar
code_scp
cluster_node_remove
code_deploy
config_diff
code_test
cluster_node_in
[www@linux-node1 scripts]$
|
測試結果
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
|
[www@linux-node1 scripts]$ tree
/deploy/
/deploy/
├── code
│ └── web-demo
│ └── index.html
├── config
│ └── web-demo
│ └── config.ini
├──
tar
└── tmp
├── web-demo_123-2017-04-23-23-12-15
│ ├── config.ini
│ └── index.html
└── web-demo_123-2017-04-23-23-13-20
├── config.ini
└── index.html
8 directories, 6 files
[www@linux-node1 scripts]$
|
版本和時間之間改成下划線
|
1
2
3
4
5
6
|
code_config(){
echo
code_config
/bin/cp
-r ${CONFIG_DIR}/* ${TMP_DIR}/
"${PRO_NAME}"
PKG_NAME=
"${PRO_NAME}"
_
"${API_VER}"
_
"${CDATE}-${CTIME}"
cd
${TMP_DIR} &&
mv
${PRO_NAME} ${PKG_NAME}
}
|
再次執行
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
|
[www@linux-node1 scripts]$ .
/deploy
.sh deploy
git pull
code_build
code_config
code_tar
code_scp
cluster_node_remove
code_deploy
config_diff
code_test
cluster_node_in
[www@linux-node1 scripts]$ tree
/deploy/
/deploy/
├── code
│ └── web-demo
│ └── index.html
├── config
│ └── web-demo
│ └── config.ini
├──
tar
└── tmp
├── web-demo_123-2017-04-23-23-12-15
│ ├── config.ini
│ └── index.html
├── web-demo_123-2017-04-23-23-13-20
│ ├── config.ini
│ └── index.html
└── web-demo_123_2017-04-23-23-17-20
├── config.ini
└── index.html
9 directories, 8 files
[www@linux-node1 scripts]$
|
可以看到tmp目錄需要定期清理
給下面2個函數加寫日志功能
|
1
2
3
4
5
6
7
8
9
10
|
code_config(){
writelog
"code_config"
/bin/cp
-r ${CONFIG_DIR}/* ${TMP_DIR}/
"${PRO_NAME}"
PKG_NAME=
"${PRO_NAME}"
_
"${API_VER}"
_
"${CDATE}-${CTIME}"
cd
${TMP_DIR} &&
mv
${PRO_NAME} ${PKG_NAME}
}
code_tar(){
writelog
"code_tar"
}
|
打包，記錄日志
|
1
2
3
4
5
|
code_tar(){
writelog
"code_tar"
cd
${TMP_DIR} &&
tar
cfz ${PKG_NAME}.
tar
.gz ${PKG_NAME}
writelog
"${PKG_NAME}.tar.gz"
}
|
再次測試腳本
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
|
[www@linux-node1 scripts]$ .
/deploy
.sh deploy
git pull
code_build
code_scp
cluster_node_remove
code_deploy
config_diff
code_test
cluster_node_in
[www@linux-node1 scripts]$ tree
/deploy/
/deploy/
├── code
│ └── web-demo
│ └── index.html
├── config
│ └── web-demo
│ └── config.ini
├──
tar
└── tmp
├── web-demo_123-2017-04-23-23-12-15
│ ├── config.ini
│ └── index.html
├── web-demo_123-2017-04-23-23-13-20
│ ├── config.ini
│ └── index.html
├── web-demo_123_2017-04-23-23-17-20
│ ├── config.ini
│ └── index.html
├── web-demo_123_2017-04-23-23-22-09
│ ├── config.ini
│ └── index.html
└── web-demo_123_2017-04-23-23-22-09.
tar
.gz
10 directories, 11 files
[www@linux-node1 scripts]$
|
准備拷貝到目標服務器
前4步都完畢，開始第五步--拷貝到目標服務器
遍歷節點
|
1
2
3
4
5
|
[www@linux-node1 scripts]$ node_list=
"192.168.58.11 192.168.58.12"
[www@linux-node1 scripts]$
for
node
in
$node_list;
do
echo
$node;
done
192.168.58.11
192.168.58.12
[www@linux-node1 scripts]$
|
腳本里添加node_list
|
1
2
3
4
5
6
7
8
9
10
11
|
#!/bin/bash
#Node List
NODE_LIST=
"192.168.58.11 192.168.58.12"
# Date/Time Veriables
LOG_DATE=
'date "+%Y-%m-%d"'
LOG_TIME=
'date "+%H-%M-%S"'
CDATE=$(
date
"+%Y-%m-%d"
)
CTIME=$(
date
"+%H-%M-%S"
)
|
分發到目標節點
|
1
2
3
4
5
6
|
code_scp(){
echo
code_scp
for
node
in
$NODE_LIST;
do
scp
${TMP_DIR}/${PKG_NAME}.
tar
.gz $node:
/opt/webroot/
done
}
|
2台機器建立webroot
|
1
2
3
4
5
6
|
[root@linux-node1 scripts]
# mkdir /opt/webroot -p
[root@linux-node1 scripts]
# chown -R www:www /opt/webroot
[root@linux-node1 scripts]
#
[root@linux-node2 ~]
# mkdir /opt/webroot -p
[root@linux-node2 ~]
# chown -R www:www /opt/webroot
[root@linux-node2 ~]
#
|
完善拷貝函數
|
1
2
3
4
5
6
|
code_scp(){
echo
code_scp
for
node
in
$NODE_LIST;
do
scp
${TMP_DIR}/${PKG_NAME}.
tar
.gz $node:
/opt/webroot/
done
}
|
再次部署測試下
|
1
2
3
4
5
6
7
8
9
10
11
12
|
[www@linux-node1 scripts]$ .
/deploy
.sh deploy
git pull
code_build
code_scp
web-demo_123_2017-04-23-23-33-50.
tar
.gz 100% 204 0.2KB
/s
00:00
web-demo_123_2017-04-23-23-33-50.
tar
.gz 100% 204 0.2KB
/s
00:00
cluster_node_remove
code_deploy
config_diff
code_test
cluster_node_in
[www@linux-node1 scripts]$
|
檢查
|
1
2
3
4
5
6
7
8
9
10
11
12
|
[www@linux-node1 scripts]$ tree
/opt/webroot/
/opt/webroot/
└── web-demo_123_2017-04-23-23-33-50.
tar
.gz
0 directories, 1
file
[www@linux-node1 scripts]$
[root@linux-node2 ~]
# tree /opt/webroot/
/opt/webroot/
└── web-demo_123_2017-04-23-23-33-50.
tar
.gz
0 directories, 1
file
[root@linux-node2 ~]
#
|
該第6步了，寫個日志代替
|
1
2
3
|
cluster_node_remove(){
writelog
"cluster_node_remove"
}
|
解壓完畢，拷貝差異文件。你要把差異文件單獨放一個目錄下。不要和配置文件放一起
修改上面，上面是相同配置目錄文件
|
1
2
3
4
5
6
|
code_config(){
writelog
"code_config"
/bin/cp
-r ${CONFIG_DIR}/* ${TMP_DIR}/
"${PRO_NAME}"
PKG_NAME=
"${PRO_NAME}"
_
"${API_VER}"
_
"${CDATE}-${CTIME}"
cd
${TMP_DIR} &&
mv
${PRO_NAME} ${PKG_NAME}
}
|
改成如下
|
1
2
3
4
5
6
|
code_config(){
writelog
"code_config"
/bin/cp
-r ${CONFIG_DIR}
/base/
* ${TMP_DIR}/
"${PRO_NAME}"
PKG_NAME=
"${PRO_NAME}"
_
"${API_VER}"
_
"${CDATE}-${CTIME}"
cd
${TMP_DIR} &&
mv
${PRO_NAME} ${PKG_NAME}
}
|
創建配置文件目錄，base存放相同的配置，other存放差異配置
|
1
2
3
4
5
6
7
8
9
|
[www@linux-node1 scripts]$
cd
/deploy/config/web-demo/
[www@linux-node1 web-demo]$
mkdir
base
[www@linux-node1 web-demo]$
mkdir
other
[www@linux-node1 web-demo]$ ll
total 4
drwxrwxr-x 2 www www 6 Apr 23 23:38 base
-rw-r--r-- 1 www www 5 Apr 23 22:46 config.ini
drwxrwxr-x 2 www www 6 Apr 23 23:38 other
[www@linux-node1 web-demo]$
|
調整下配置文件所在目錄
|
1
2
3
4
5
6
7
|
[www@linux-node1 web-demo]$
mv
config.ini base/
[www@linux-node1 web-demo]$
cd
other/
[www@linux-node1 other]$
echo
192.168.58.12-config >>192.168.58.12.
crontab
.xml
[www@linux-node1 other]$ ll
total 4
-rw-rw-r-- 1 www www 21 Apr 23 23:39 192.168.58.12.
crontab
.xml
[www@linux-node1 other]$
|
拷貝差異文件到目標服務器的目標目錄
|
1
2
3
4
5
6
7
8
9
|
code_deploy(){
echo
code_deploy
cd
/opt/webroot/
&&
tar
xfz ${PKG_NAME}.
tar
.gz
}
config_diff(){
echo
config_diff
scp
${CONFIG_DIR}
/other/192
.168.58.12.
crontab
.xml 192.168.58.12:
/opt/webroot/
${PKG_NAME}
}
|
再次測試
|
1
2
3
4
5
6
7
8
9
10
11
12
|
[www@linux-node1 scripts]$ .
/deploy
.sh deploy
git pull
code_build
code_scp
web-demo_123_2017-04-23-23-43-48.
tar
.gz 100% 204 0.2KB
/s
00:00
web-demo_123_2017-04-23-23-43-48.
tar
.gz 100% 204 0.2KB
/s
00:00
code_deploy
config_diff
192.168.58.12.
crontab
.xml 100% 21 0.0KB
/s
00:00
code_test
cluster_node_in
[www@linux-node1 scripts]$
|
上面還有不足的地方，scp到目標服務器並解壓，應該使用ssh遠程執行、。上面腳本遠程node2上解壓是失敗的
腳本再次改造下，把部署的函數和差異配置合並到一起
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
|
cluster_node_remove(){
writelog
"cluster_node_remove"
}
code_deploy(){
echo
code_deploy
for
node
in
$NODE_LIST;
do
ssh
$node
"cd /opt/webroot/ && tar xfz ${PKG_NAME}.tar.gz"
done
scp
${CONFIG_DIR}
/other/192
.168.58.12.
crontab
.xml 192.168.58.12:
/opt/webroot/
${PKG_NAME}
/crontab
.xml
}
code_test(){
echo
code_test
}
|
創建webroot
|
1
2
3
4
5
6
7
8
|
[root@linux-node1 ~]
# mkdir /webroot
[root@linux-node1 ~]
# chown -R www:www /webroot
[root@linux-node1 ~]
#
[root@linux-node2 ~]
# mkdir /webroot
[root@linux-node2 ~]
# chown -R www:www /webroot
[root@linux-node2 ~]
#
|
再次對部署函數優化，添加使用軟鏈接參數（這個是秒級回滾的關鍵）
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
|
cluster_node_remove(){
writelog
"cluster_node_remove"
}
code_deploy(){
echo
code_deploy
for
node
in
$NODE_LIST;
do
ssh
$node
"cd /opt/webroot/ && tar xfz ${PKG_NAME}.tar.gz"
done
scp
${CONFIG_DIR}
/other/192
.168.58.12.
crontab
.xml 192.168.58.12:
/opt/webroot/
${PKG_NAME}
/crontab
.xml
ln
-s
/opt/webroot/
${PKG_NAME}
/webroot/web-demo
}
code_test(){
echo
code_test
}
|
再次對腳本優化
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
|
cluster_node_remove(){
writelog
"cluster_node_remove"
}
code_deploy(){
echo
code_deploy
for
node
in
$NODE_LIST;
do
ssh
$node
"cd /opt/webroot/ && tar xfz ${PKG_NAME}.tar.gz"
done
scp
${CONFIG_DIR}
/other/192
.168.58.12.
crontab
.xml 192.168.58.12:
/opt/webroot/
${PKG_NAME}
/crontab
.xml
rm
-f
/webroot/web-demo
&&
ln
-s
/opt/webroot/
${PKG_NAME}
/webroot/web-demo
}
code_test(){
echo
code_test
}
|
自動化部署的精髓，創建軟鏈接
對腳本優化，每個服務器要執行相同的操作，因此放在循環里
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
|
cluster_node_remove(){
writelog
"cluster_node_remove"
}
code_deploy(){
echo
code_deploy
for
node
in
$NODE_LIST;
do
ssh
$node
"cd /opt/webroot/ && tar xfz ${PKG_NAME}.tar.gz"
rm
-f
/webroot/web-demo
&&
ln
-s
/opt/webroot/
${PKG_NAME}
/webroot/web-demo
done
scp
${CONFIG_DIR}
/other/192
.168.58.12.
crontab
.xml 192.168.58.12:
/opt/webroot/
${PKG_NAME}
/crontab
.xml
}
code_test(){
echo
code_test
}
|
拷貝差異文件應該創建軟鏈接之后拷貝，其實就是路徑寫的少點
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
|
cluster_node_remove(){
writelog
"cluster_node_remove"
}
code_deploy(){
echo
code_deploy
for
node
in
$NODE_LIST;
do
ssh
$node
"cd /opt/webroot/ && tar xfz ${PKG_NAME}.tar.gz"
rm
-f
/webroot/web-demo
&&
ln
-s
/opt/webroot/
${PKG_NAME}
/webroot/web-demo
done
scp
${CONFIG_DIR}
/other/192
.168.58.12.
crontab
.xml 192.168.58.12:
/webroot/web-demo/crontab
.xml
}
code_test(){
echo
code_test
}
|
第一次沒軟鏈接，會報錯。需要先手動創建個軟連接，或者先創建web-demo目錄
其實我覺的完全可以在上面腳本里加入mkdir /webroot/web-demo -p 這樣永遠不會錯
|
1
2
3
4
5
|
[www@linux-node1 scripts]$
cd
/webroot/
[www@linux-node1 webroot]$
mkdir
web-demo -p
[www@linux-node1 webroot]$
[root@linux-node2 webroot]
# mkdir web-demo -p
[root@linux-node2 webroot]
#
|
再次執行部署腳本。node1完成了
繼續優化腳本，rm -rf 這里要寫為-rf
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
|
cluster_node_remove(){
writelog
"cluster_node_remove"
}
code_deploy(){
echo
code_deploy
for
node
in
$NODE_LIST;
do
ssh
$node
"cd /opt/webroot/ && tar xfz ${PKG_NAME}.tar.gz"
ssh
$node
"rm -rf /webroot/web-demo && ln -s /opt/webroot/${PKG_NAME} /webroot/web-demo"
done
scp
${CONFIG_DIR}
/other/192
.168.58.12.
crontab
.xml 192.168.58.12:
/webroot/web-demo/crontab
.xml
}
code_test(){
echo
code_test
}
|
測試腳本
|
1
2
3
4
5
6
7
8
9
10
11
12
|
[www@linux-node1 scripts]$ .
/deploy
.sh deploy
git pull
code_build
code_scp
web-demo_123_2017-04-24-00-01-24.
tar
.gz 100% 204 0.2KB
/s
00:00
web-demo_123_2017-04-24-00-01-24.
tar
.gz 100% 204 0.2KB
/s
00:00
code_deploy
192.168.58.12.
crontab
.xml 100% 21 0.0KB
/s
00:00
.
/deploy
.sh: line 113: config_diff:
command
not found
code_test
cluster_node_in
[www@linux-node1 scripts]$
|
主函數里，刪除這個 config_diff
繼續測試和檢查
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
|
測試
[www@linux-node1 scripts]$ .
/deploy
.sh deploy
git pull
code_build
code_scp
web-demo_123_2017-04-24-00-02-44.
tar
.gz 100% 205 0.2KB
/s
00:00
web-demo_123_2017-04-24-00-02-44.
tar
.gz 100% 205 0.2KB
/s
00:00
code_deploy
192.168.58.12.
crontab
.xml 100% 21 0.0KB
/s
00:00
code_test
cluster_node_in
[www@linux-node1 scripts]$
檢查
[www@linux-node1 scripts]$ ll
/webroot/
total 0
lrwxrwxrwx 1 www www 45 Apr 24 00:02 web-demo ->
/opt/webroot/web-demo_123_2017-04-24-00-02-44
[www@linux-node1 scripts]$
[root@linux-node2 webroot]
# ll /webroot/
total 0
lrwxrwxrwx 1 www www 45 Apr 24 00:02 web-demo ->
/opt/webroot/web-demo_123_2017-04-24-00-02-44
[root@linux-node2 webroot]
#
|
模擬版本更新，把版本寫成456
|
1
2
3
4
5
6
|
code_get(){
writelog
"code_get"
;
cd
$CODE_DIR &&
echo
"git pull"
cp
-r ${CODE_DIR} ${TMP_DIR}/
API_VER=
"456"
}
|
繼續測試
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
|
[www@linux-node1 scripts]$ .
/deploy
.sh deploy
git pull
code_build
code_scp
web-demo_456_2017-04-24-00-04-05.
tar
.gz 100% 204 0.2KB
/s
00:00
web-demo_456_2017-04-24-00-04-05.
tar
.gz 100% 204 0.2KB
/s
00:00
code_deploy
192.168.58.12.
crontab
.xml 100% 21 0.0KB
/s
00:00
code_test
cluster_node_in
[www@linux-node1 scripts]$ ll
/webroot/
total 0
lrwxrwxrwx 1 www www 45 Apr 24 00:04 web-demo ->
/opt/webroot/web-demo_456_2017-04-24-00-04-05
[www@linux-node1 scripts]$
|
檢查
鏈接 到了新的版本
#################################################################################
node1和node2都裝apache
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
|
[root@linux-node1 ~]
# yum install httpd -y
Loaded plugins: fastestmirror
Loading mirror speeds from cached hostfile
* base: mirrors.zju.edu.cn
* epel: mirrors.tuna.tsinghua.edu.cn
* extras: mirrors.zju.edu.cn
* updates: mirrors.163.com
Package httpd-2.4.6-45.el7.centos.4.x86_64 already installed and latest version
Nothing to
do
[root@linux-node1 ~]
#
[root@linux-node2 ~]
# yum install httpd -y
Loaded plugins: fastestmirror
Loading mirror speeds from cached hostfile
* base: mirrors.zju.edu.cn
* extras: mirrors.zju.edu.cn
* updates: mirrors.163.com
Package httpd-2.4.6-45.el7.centos.4.x86_64 already installed and latest version
Nothing to
do
[root@linux-node2 ~]
#
|
把apache的根目錄都改成/opt/webroot
|
1
2
|
[root@linux-node2 ~]
# vim /etc/httpd/conf/httpd.conf
[root@linux-node2 ~]
#
|
上面是默認的，改成如下配置
node1也改成如下
啟動apache
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
|
[root@linux-node2 ~]
# systemctl start httpd
[root@linux-node2 ~]
# lsof -i:80
COMMAND PID USER FD TYPE DEVICE SIZE
/OFF
NODE NAME
httpd 21937 root 4u IPv6 132900 0t0 TCP *:http (LISTEN)
httpd 21938 apache 4u IPv6 132900 0t0 TCP *:http (LISTEN)
httpd 21939 apache 4u IPv6 132900 0t0 TCP *:http (LISTEN)
httpd 21940 apache 4u IPv6 132900 0t0 TCP *:http (LISTEN)
httpd 21941 apache 4u IPv6 132900 0t0 TCP *:http (LISTEN)
httpd 21942 apache 4u IPv6 132900 0t0 TCP *:http (LISTEN)
[root@linux-node2 ~]
#
[root@linux-node1 ~]
# systemctl start httpd
[root@linux-node1 ~]
# lsof -i:80
COMMAND PID USER FD TYPE DEVICE SIZE
/OFF
NODE NAME
httpd 23765 root 4u IPv6 117035 0t0 TCP *:http (LISTEN)
httpd 23767 apache 4u IPv6 117035 0t0 TCP *:http (LISTEN)
httpd 23768 apache 4u IPv6 117035 0t0 TCP *:http (LISTEN)
httpd 23769 apache 4u IPv6 117035 0t0 TCP *:http (LISTEN)
httpd 23770 apache 4u IPv6 117035 0t0 TCP *:http (LISTEN)
httpd 23771 apache 4u IPv6 117035 0t0 TCP *:http (LISTEN)
[root@linux-node1 ~]
#
|
訪問網頁
修改首頁
|
1
2
3
4
5
6
7
8
9
10
11
|
[www@linux-node1 ~]$
cd
/deploy/code/
[www@linux-node1 code]$
cd
web-demo/
[www@linux-node1 web-demo]$
pwd
/deploy/code/web-demo
[www@linux-node1 web-demo]$ ll
total 4
-rw-r--r-- 1 www www 9 Apr 26 22:28 index.html
[www@linux-node1 web-demo]$
cat
index.html
nmap.com
[www@linux-node1 web-demo]$
cd
/scripts/
[www@linux-node1 scripts]$
|
再次部署
|
1
2
3
4
5
6
7
8
9
10
11
|
[www@linux-node1 scripts]$ .
/deploy
.sh deploy
git pull
code_build
code_scp
web-demo_456_2017-04-27-21-40-36.
tar
.gz 100% 220 0.2KB
/s
00:00
web-demo_456_2017-04-27-21-40-36.
tar
.gz 100% 220 0.2KB
/s
00:00
code_deploy
192.168.58.12.
crontab
.xml 100% 21 0.0KB
/s
00:00
code_test
cluster_node_in
[www@linux-node1 scripts]$
|
檢查部署結果，注意ctrl+f5
查看此時自動化部署腳本的配置
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
|
[root@linux-node1 scripts]
# cat deploy.sh
#!/bin/bash
#Dir List
#mkdir -p /deploy/code/web-demo/
#mkdir -p /deploy/config/web-demo/base
#mkdir -p /deploy/config/web-demo/other
#mkdir -p /deploy/tar
#mkdir -p /deploy/tmp
#mkdir -p /opt/webroot
#mkdir -p /webroot
#chown -R www:www /deploy
#chown -R www:www /opt/webroot
#chown -R www:www /webroot
#Node List
NODE_LIST=
"192.168.58.11 192.168.58.12"
# Date/Time Veriables
LOG_DATE=
'date "+%Y-%m-%d"'
LOG_TIME=
'date "+%H-%M-%S"'
CDATE=$(
date
"+%Y-%m-%d"
)
CTIME=$(
date
"+%H-%M-%S"
)
#Shell Env
SHELL_NAME=
"deploy.sh"
SHELL_DIR=
"/home/www"
SHELL_LOG=
"${SHELL_DIR}/${SHELL_NAME}.log"
#Code Env
PRO_NAME=
"web-demo"
CODE_DIR=
"/deploy/code/web-demo"
CONFIG_DIR=
"/deploy/config/web-demo"
TMP_DIR=
"/deploy/tmp"
TAR_DIR=
"/deploy/tar"
LOCK_FILE=
"/tmp/deploy.lock"
usage(){
echo
$
"Usage: $0 [ deploy | rollback ]"
}
writelog(){
LOGINFO=$1
echo
"${CDATE} ${CTIME}: ${SHELL_NAME} : ${LOGINFO}"
>> ${SHELL_LOG}
}
shell_lock(){
touch
${LOCK_FILE}
}
shell_unlock(){
rm
-f ${LOCK_FILE}
}
code_get(){
writelog
"code_get"
;
cd
$CODE_DIR &&
echo
"git pull"
cp
-r ${CODE_DIR} ${TMP_DIR}/
API_VER=
"456"
}
code_build(){
echo
code_build
}
code_config(){
writelog
"code_config"
/bin/cp
-r ${CONFIG_DIR}
/base/
* ${TMP_DIR}/
"${PRO_NAME}"
PKG_NAME=
"${PRO_NAME}"
_
"${API_VER}"
_
"${CDATE}-${CTIME}"
cd
${TMP_DIR} &&
mv
${PRO_NAME} ${PKG_NAME}
}
code_tar(){
writelog
"code_tar"
cd
${TMP_DIR} &&
tar
cfz ${PKG_NAME}.
tar
.gz ${PKG_NAME}
writelog
"${PKG_NAME}.tar.gz"
}
code_scp(){
echo
"code_scp"
for
node
in
$NODE_LIST;
do
scp
${TMP_DIR}/${PKG_NAME}.
tar
.gz $node:
/opt/webroot/
done
}
cluster_node_remove(){
writelog
"cluster_node_remove"
}
code_deploy(){
echo
code_deploy
for
node
in
$NODE_LIST;
do
ssh
$node
"cd /opt/webroot/ && tar xfz ${PKG_NAME}.tar.gz"
ssh
$node
"rm -rf /webroot/web-demo && ln -s /opt/webroot/${PKG_NAME} /webroot/web-demo"
done
scp
${CONFIG_DIR}
/other/192
.168.58.12.
crontab
.xml 192.168.58.12:
/webroot/web-demo/crontab
.xml
}
code_test(){
echo
code_test
}
cluster_node_in(){
echo
cluster_node_in
}
rollback(){
echo
rollback
}
main(){
if
[ -f ${LOCK_FILE} ];
then
echo
"Deploy is running"
&&
exit
;
fi
DEPLOY_METHOD=$1
case
$DEPLOY_METHOD
in
deploy)
shell_lock;
code_get;
code_build;
code_config;
code_tar;
code_scp;
cluster_node_remove;
code_deploy;
code_test;
cluster_node_in;
shell_unlock;
;;
rollback)
shell_lock;
rollback;
shell_unlock;
;;
*)
usage;
esac
}
main $1
[root@linux-node1 scripts]
#
|
修改測試部分,完善測試函數
過濾到，返回0，過濾不到返回其它的
-s靜默模式
|
1
2
3
4
5
6
7
8
|
[root@linux-node1 scripts]
# curl --head http://192.168.58.11/index.html | grep "200 OK"
% Total % Received % Xferd Average Speed Time Time Time Current
Dload Upload Total Spent Left Speed
0 9 0 0 0 0 0 0 --:--:-- --:--:-- --:--:-- 0
HTTP
/1
.1 200 OK
[root@linux-node1 scripts]
# curl -s --head http://192.168.58.11/index.html | grep "200 OK"
HTTP
/1
.1 200 OK
[root@linux-node1 scripts]
#
|
准備是這樣的，部署一個，測試一個，通了才繼續往下部署。這樣才合理
很多公司拿預熱節點部署。通過之后再部署生產節點
先修改下上面，把主機分組，模擬分組部署
GROUP1_LIST模擬預熱節點
GROUP2_LIST 模擬其余的全部節點
|
1
2
3
4
5
6
7
8
9
10
|
#chown -R www:www /opt/webroot
#chown -R www:www /webroot
#Node List
GROUP1_LIST=
"192.168.58.11"
GROUP2_LIST=
"192.168.58.12"
# Date/Time Veriables
LOG_DATE=
'date "+%Y-%m-%d"'
LOG_TIME=
'date "+%H-%M-%S"'
|
scp這里改下
|
1
2
3
4
5
6
7
8
9
|
code_scp(){
writelog
"code_scp"
for
node
in
$GROUP1_LIST;
do
scp
${TMP_DIR}/${PKG_NAME}.
tar
.gz $node:
/opt/webroot/
done
for
node
in
$GROUP2_LIST;
do
scp
${TMP_DIR}/${PKG_NAME}.
tar
.gz $node:
/opt/webroot/
done
}
|
把code_deploy也修改下
下面這里是修改之前
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
|
cluster_node_remove(){
writelog
"cluster_node_remove"
}
code_deploy(){
echo
code_deploy
for
node
in
$NODE_LIST;
do
ssh
$node
"cd /opt/webroot/ && tar xfz ${PKG_NAME}.tar.gz"
ssh
$node
"rm -rf /webroot/web-demo && ln -s /opt/webroot/${PKG_NAME} /webroot/web-demo"
done
scp
${CONFIG_DIR}
/other/192
.168.58.12.
crontab
.xml 192.168.58.12:
/webroot/web-demo/crontab
.xml
}
code_test(){
echo
code_test
}
|
修改之后
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
|
cluster_node_remove(){
writelog
"cluster_node_remove"
}
group1_deploy(){
echo
code_deploy
for
node
in
$GROUP1_LIST;
do
ssh
$node
"cd /opt/webroot/ && tar xfz ${PKG_NAME}.tar.gz"
ssh
$node
"rm -rf /webroot/web-demo && ln -s /opt/webroot/${PKG_NAME} /webroot/web-demo"
done
}
group2_deploy(){
echo
code_deploy
for
node
in
$GROUP2_LIST;
do
ssh
$node
"cd /opt/webroot/ && tar xfz ${PKG_NAME}.tar.gz"
ssh
$node
"rm -rf /webroot/web-demo && ln -s /opt/webroot/${PKG_NAME} /webroot/web-demo"
done
scp
${CONFIG_DIR}
/other/192
.168.58.12.
crontab
.xml 192.168.58.12:
/webroot/web-demo/crontab
.xml
}
cluster_node_in(){
echo
cluster_node_in
}
|
第一組弄的只有1台機器，作為預生產節點
一個節點如果重啟需要1分鍾的話，5個節點豈不是5分鍾。
所以，1個節點先作為預生產節點，部署完畢，測試完畢之后，其它節點可以直接部署了。思想上是這樣
添加group1_test測試函數
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
|
group1_deploy(){
echo
code_deploy
for
node
in
$GROUP1_LIST;
do
ssh
$node
"cd /opt/webroot/ && tar xfz ${PKG_NAME}.tar.gz"
ssh
$node
"rm -rf /webroot/web-demo && ln -s /opt/webroot/${PKG_NAME} /webroot/web-demo"
done
}
group1_test(){
curl -s --
head
http:
//192
.168.58.11
/index
.html |
grep
"200 OK"
if
[ $? -
ne
0 ];
then
shell_unlock;
writelog
"test error"
&&
exit
;
fi
}
group2_deploy(){
echo
code_deploy
for
node
in
$GROUP2_LIST;
do
ssh
$node
"cd /opt/webroot/ && tar xfz ${PKG_NAME}.tar.gz"
ssh
$node
"rm -rf /webroot/web-demo && ln -s /opt/webroot/${PKG_NAME} /webroot/web-demo"
done
scp
${CONFIG_DIR}
/other/192
.168.58.12.
crontab
.xml 192.168.58.12:
/webroot/web-demo/crontab
.xml
}
|
由於group2也要測試，並且每個節點都要測試
因此可以把測試的部分提取出來，封裝成函數
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
|
shell_lock(){
touch
${LOCK_FILE}
}
url_test(){
URL=$1
curl -s --
head
$URL |
grep
"200 OK"
if
[ $? -
ne
0 ];
then
shell_unlock;
writelog
"test error"
&&
exit
;
fi
}
shell_unlock(){
rm
-f ${LOCK_FILE}
}
|
測試部分可以這么寫
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
|
cluster_node_remove(){
writelog
"cluster_node_remove"
}
group1_deploy(){
echo
“code_deploy”
for
node
in
$GROUP1_LIST;
do
ssh
$node
"cd /opt/webroot/ && tar xfz ${PKG_NAME}.tar.gz"
ssh
$node
"rm -rf /webroot/web-demo && ln -s /opt/webroot/${PKG_NAME} /webroot/web-demo"
done
}
group1_test(){
curl_test
"http://192.168.58.11/index.html"
echo
"add to cluster"
for
node
in
$GROUP1_LIST;
do
ssh
$node
"cd /opt/webroot/ && tar xfz ${PKG_NAME}.tar.gz"
ssh
$node
"rm -rf /webroot/web-demo && ln -s /opt/webroot/${PKG_NAME} /webroot/web-demo"
done
}
group1_test(){
url_test
"http://192.168.58.11/index.html"
echo
"add to cluster"
}
group2_deploy(){
echo
“code_deploy”
for
node
in
$GROUP2_LIST;
do
ssh
$node
"cd /opt/webroot/ && tar xfz ${PKG_NAME}.tar.gz"
ssh
$node
"rm -rf /webroot/web-demo && ln -s /opt/webroot/${PKG_NAME} /webroot/web-demo"
done
scp
${CONFIG_DIR}
/other/192
.168.58.12.
crontab
.xml 192.168.58.12:
/webroot/web-demo/crontab
.xml
}
group2_test(){
url_test
"http://192.168.58.12/index.html"
echo
"add to cluster"
}
cluster_node_in(){
echo
cluster_node_in
}
|
可以把下面函數刪除。是否可以加入集群，放在測試函數即可
|
1
2
3
|
cluster_node_in(){
echo
cluster_node_in
}
|
主函數會這么寫
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
|
cluster_node_remove(){
writelog
"cluster_node_remove"
DEPLOY_METHOD=$1
case
$DEPLOY_METHOD
in
deploy)
shell_lock;
code_get;
code_build;
code_config;
code_tar;
code_scp;
group1_deploy;
group1_test;
group2_deploy;
group2_test;
shell_unlock;
;;
rollback)
shell_lock;
rollback;
shell_unlock;
;;
*)
usage;
esac
}
main $1
|
測試腳本
|
1
2
3
4
5
6
7
8
9
10
11
12
13
|
[www@linux-node1 scripts]$ .
/deploy
.sh deploy
git pull
code_build
web-demo_456_2017-04-27-22-41-54.
tar
.gz 100% 214 0.2KB
/s
00:00
web-demo_456_2017-04-27-22-41-54.
tar
.gz 100% 214 0.2KB
/s
00:00
code_deploy
HTTP
/1
.1 200 OK
add to cluster
code_deploy
192.168.58.12.
crontab
.xml 100% 21 0.0KB
/s
00:00
HTTP
/1
.1 200 OK
add to cluster
[www@linux-node1 scripts]$
|
沒問題，可以訪問
繼續優化腳本
pre是預熱節點，生產只有1個
group1_list里面有多個
|
1
2
3
4
5
6
7
8
9
10
|
#Node List
PRE_LIST=
"192.168.58.11"
GROUP1_LIST=
"192.168.58.12"
# Date/Time Veriables
LOG_DATE=
'date "+%Y-%m-%d"'
LOG_TIME=
'date "+%H-%M-%S"'
CDATE=$(
date
"+%Y-%m-%d"
)
CTIME=$(
date
"+%H-%M-%S"
)
|
下面這里改下
|
1
2
3
4
5
6
7
8
9
|
code_scp(){
writelog
"code_scp"
for
node
in
$PRE_LIST;
do
scp
${TMP_DIR}/${PKG_NAME}.
tar
.gz $node:
/opt/webroot/
done
for
node
in
$GROUP1_LIST;
do
scp
${TMP_DIR}/${PKG_NAME}.
tar
.gz $node:
/opt/webroot/
done
}
|
上面的remove函數不需要了。刪除即可，如下
|
1
2
3
|
cluster_node_remove(){
writelog
"cluster_node_remove"
}
|
下面的group_list 名字還需要改，再次優化下，把原先的group1_list改成pre_test，然后下面這里的改成group1
下面的group1_test應該改成for循環，測試里面的節點，可以測一個加一個節點到集群。還可以測試完，一塊加測試的時候，
如果你只有2-3個節點，沒必要寫for循環
優化之后內容
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
|
[root@linux-node1 scripts]
# sed -n '89,128p' deploy.sh
code_scp(){
writelog
"code_scp"
for
node
in
$PRE_LIST;
do
scp
${TMP_DIR}/${PKG_NAME}.
tar
.gz $node:
/opt/webroot/
done
for
node
in
$GROUP1_LIST;
do
scp
${TMP_DIR}/${PKG_NAME}.
tar
.gz $node:
/opt/webroot/
done
}
pre_deploy(){
writelog
"remove from cluster"
ssh
$PRE_LIST
"cd /opt/webroot/ && tar xfz ${PKG_NAME}.tar.gz"
ssh
$PRE_LIST
"rm -rf /webroot/web-demo && ln -s /opt/webroot/${PKG_NAME} /webroot/web-demo"
}
pre_test(){
url_test
"http://${PRE_LIST}/index.html"
echo
"add to cluster"
}
group1_deploy(){
writelog
"remove from cluster"
for
node
in
$GROUP1_LIST;
do
ssh
$node
"cd /opt/webroot/ && tar xfz ${PKG_NAME}.tar.gz"
ssh
$node
"rm -rf /webroot/web-demo && ln -s /opt/webroot/${PKG_NAME} /webroot/web-demo"
done
scp
${CONFIG_DIR}
/other/192
.168.58.12.
crontab
.xml 192.168.58.12:
/webroot/web-demo/crontab
.xml
}
group1_test(){
url_test
"http://192.168.58.12/index.html"
echo
"add to cluster"
}
rollback(){
echo
rollback
}
[root@linux-node1 scripts]
#
|
測試的時候，如果只有幾個節點，直接幾行地址就行了嘛。沒必要寫太復雜
|
1
2
3
4
5
6
|
group1_test(){
url_test
"http://192.168.58.12/index.html"
url_test
"http://192.168.58.12/index.html"
url_test
"http://192.168.58.12/index.html"
echo
"add to cluster"
}
|
主函數部分改成下面
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
|
[root@linux-node1 scripts]
# sed -n '130,$p' deploy.sh
main(){
if
[ -f ${LOCK_FILE} ];
then
echo
"Deploy is running"
&&
exit
;
fi
DEPLOY_METHOD=$1
case
$DEPLOY_METHOD
in
deploy)
shell_lock;
code_get;
code_build;
code_config;
code_tar;
code_scp;
pre_deploy;
pre_test;
group1_deploy;
group1_test;
shell_unlock;
;;
rollback)
shell_lock;
rollback;
shell_unlock;
;;
*)
usage;
esac
}
main $1
[root@linux-node1 scripts]
#
|
改下代碼，區分下主頁顯示和之前的不同
|
1
2
3
4
5
6
7
|
[www@linux-node1 scripts]$
cd
/deploy/code/web-demo/
[www@linux-node1 web-demo]$
cat
index.html
nmap.com
[www@linux-node1 web-demo]$
echo
"www.nmap.com"
> index.html
[www@linux-node1 web-demo]$
cat
index.html
www.nmap.com
[www@linux-node1 web-demo]$
|
再次部署
|
1
2
3
4
5
6
7
8
9
10
11
|
[www@linux-node1 scripts]$ .
/deploy
.sh deploy
git pull
code_build
web-demo_456_2017-04-27-23-17-27.
tar
.gz 100% 220 0.2KB
/s
00:00
web-demo_456_2017-04-27-23-17-27.
tar
.gz 100% 220 0.2KB
/s
00:00
HTTP
/1
.1 200 OK
add to cluster
192.168.58.12.
crontab
.xml 100% 21 0.0KB
/s
00:00
HTTP
/1
.1 200 OK
add to cluster
[www@linux-node1 scripts]$
|
檢查頁面，沒問題
此時腳本內容如下
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
|
[root@linux-node1 scripts]
# cat deploy.sh
#!/bin/bash
#Dir List
#mkdir -p /deploy/code/web-demo/
#mkdir -p /deploy/config/web-demo/base
#mkdir -p /deploy/config/web-demo/other
#mkdir -p /deploy/tar
#mkdir -p /deploy/tmp
#mkdir -p /opt/webroot
#mkdir -p /webroot
#chown -R www:www /deploy
#chown -R www:www /opt/webroot
#chown -R www:www /webroot
#Node List
PRE_LIST=
"192.168.58.11"
GROUP1_LIST=
"192.168.58.12"
# Date/Time Veriables
LOG_DATE=
'date "+%Y-%m-%d"'
LOG_TIME=
'date "+%H-%M-%S"'
CDATE=$(
date
"+%Y-%m-%d"
)
CTIME=$(
date
"+%H-%M-%S"
)
#Shell Env
SHELL_NAME=
"deploy.sh"
SHELL_DIR=
"/home/www"
SHELL_LOG=
"${SHELL_DIR}/${SHELL_NAME}.log"
#Code Env
PRO_NAME=
"web-demo"
CODE_DIR=
"/deploy/code/web-demo"
CONFIG_DIR=
"/deploy/config/web-demo"
TMP_DIR=
"/deploy/tmp"
TAR_DIR=
"/deploy/tar"
LOCK_FILE=
"/tmp/deploy.lock"
usage(){
echo
$
"Usage: $0 [ deploy | rollback ]"
}
writelog(){
LOGINFO=$1
echo
"${CDATE} ${CTIME}: ${SHELL_NAME} : ${LOGINFO}"
>> ${SHELL_LOG}
}
shell_lock(){
touch
${LOCK_FILE}
}
url_test(){
URL=$1
curl -s --
head
$URL |
grep
"200 OK"
if
[ $? -
ne
0 ];
then
shell_unlock;
writelog
"test error"
&&
exit
;
fi
}
shell_unlock(){
rm
-f ${LOCK_FILE}
}
code_get(){
writelog
"code_get"
;
cd
$CODE_DIR &&
echo
"git pull"
cp
-r ${CODE_DIR} ${TMP_DIR}/
API_VER=
"456"
}
code_build(){
echo
code_build
}
code_config(){
writelog
"code_config"
/bin/cp
-r ${CONFIG_DIR}
/base/
* ${TMP_DIR}/
"${PRO_NAME}"
PKG_NAME=
"${PRO_NAME}"
_
"${API_VER}"
_
"${CDATE}-${CTIME}"
cd
${TMP_DIR} &&
mv
${PRO_NAME} ${PKG_NAME}
}
code_tar(){
writelog
"code_tar"
cd
${TMP_DIR} &&
tar
cfz ${PKG_NAME}.
tar
.gz ${PKG_NAME}
writelog
"${PKG_NAME}.tar.gz"
}
code_scp(){
writelog
"code_scp"
for
node
in
$PRE_LIST;
do
scp
${TMP_DIR}/${PKG_NAME}.
tar
.gz $node:
/opt/webroot/
done
for
node
in
$GROUP1_LIST;
do
scp
${TMP_DIR}/${PKG_NAME}.
tar
.gz $node:
/opt/webroot/
done
}
pre_deploy(){
writelog
"remove from cluster"
ssh
$PRE_LIST
"cd /opt/webroot/ && tar xfz ${PKG_NAME}.tar.gz"
ssh
$PRE_LIST
"rm -rf /webroot/web-demo && ln -s /opt/webroot/${PKG_NAME} /webroot/web-demo"
}
pre_test(){
url_test
"http://${PRE_LIST}/index.html"
echo
"add to cluster"
}
group1_deploy(){
writelog
"remove from cluster"
for
node
in
$GROUP1_LIST;
do
ssh
$node
"cd /opt/webroot/ && tar xfz ${PKG_NAME}.tar.gz"
ssh
$node
"rm -rf /webroot/web-demo && ln -s /opt/webroot/${PKG_NAME} /webroot/web-demo"
done
scp
${CONFIG_DIR}
/other/192
.168.58.12.
crontab
.xml 192.168.58.12:
/webroot/web-demo/crontab
.xml
}
group1_test(){
url_test
"http://192.168.58.12/index.html"
echo
"add to cluster"
}
rollback(){
echo
rollback
}
main(){
if
[ -f ${LOCK_FILE} ];
then
echo
"Deploy is running"
&&
exit
;
fi
DEPLOY_METHOD=$1
case
$DEPLOY_METHOD
in
deploy)
shell_lock;
code_get;
code_build;
code_config;
code_tar;
code_scp;
pre_deploy;
pre_test;
group1_deploy;
group1_test;
shell_unlock;
;;
rollback)
shell_lock;
rollback;
shell_unlock;
;;
*)
usage;
esac
}
main $1
[root@linux-node1 scripts]
#
|
自動化部署實戰-秒級回滾
再加個能緊急回滾的，就不考慮用戶體驗了。以緊急回滾為主。也不測試了
更緊急流程。直接執行回滾上一個版本
先寫個第一種正常流程的，列出回滾版本的函數
因此ROLLBACK_VER=$2 和rollback $ROLLBACK_VER ;
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
|
[root@linux-node1 scripts]
# sed -n '129,$p' deploy.sh
main(){
if
[ -f ${LOCK_FILE} ];
then
echo
"Deploy is running"
&&
exit
;
fi
DEPLOY_METHOD=$1
ROLLBACK_VER=$2
case
$DEPLOY_METHOD
in
deploy)
shell_lock;
code_get;
code_build;
code_config;
code_tar;
code_scp;
pre_deploy;
pre_test;
group1_deploy;
group1_test;
shell_unlock;
;;
rollback)
shell_lock;
rollback $ROLLBACK_VER ;
shell_unlock;
;;
*)
usage;
esac
}
main $1
[root@linux-node1 scripts]
#
|
完善rollback函數
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
|
group1_test(){
url_test
"http://192.168.58.12/index.html"
echo
"add to cluster"
}
rollback_fun(){
for
node
in
$ROLLBACK_LIST;
do
ssh
$node
"rm -f /webroot/web-demo && ln-s /opt/webroo/$1 /webroot/web-demo"
done
}
rollback(){
case
$1
in
list)
ls
-l
/opt/webroot/
*.
tar
.gz
;;
*)
rollback_fun $1
esac
}
main(){
if
[ -f ${LOCK_FILE} ];
then
|
最前面加上如下行，然后寫上所有節點
|
1
2
3
4
5
6
7
8
|
#Node List
PRE_LIST=
"192.168.58.11"
GROUP1_LIST=
"192.168.58.12"
ROLLBACK_LIST=
"192.168.58.11 192.168.58.12"
# Date/Time Veriables
LOG_DATE=
'date "+%Y-%m-%d"'
LOG_TIME=
'date "+%H-%M-%S"'
|
下面這里改成如下
|
1
2
3
|
usage(){
echo
$
"Usage: $0 { deploy | rollback [ list |version ]}"
}
|
上面寫不合適，應該不傳參數列出來可以回退的版本
現在如果rollback一個不存在的版本，它會把軟鏈接刪除了，回退也失敗，因此可以反過來
再改下
之所以把這個if判斷寫在for下面。主要是，如果只部署了預生產節點，沒部署其它節點
那么回退的時候就可以把預部署節點回退了。否則其它的都回退失敗了，並且還要檢測其它節點上有沒有包。我覺得這里應該ssh -d檢測
|
1
2
3
4
5
6
7
|
rollback_fun(){
for
node
in
$ROLLBACK_LIST;
do
if
[ -d
/opt/webroot/
$1 ];
then
ssh
$node
"rm -f /webroot/web-demo && ln -s /opt/webroo/$1 /webroot/web-demo"
fi
done
}
|
繼續改造（其實這里通過ssh之后判斷其余節點目錄是否存在的）
|
1
2
3
4
5
6
7
|
rollback_fun(){
for
node
in
$ROLLBACK_LIST;
do
if
[ -d
/opt/webroot/
$1 ];
then
ssh
$node
"[ -d /opt/webroot/$1 ] && rm -f /webroot/web-demo && ln -s /opt/webroo/$1 /webroot/web-demo"
fi
done
}
|
這里就不對了吧。rollback之后沒跟參數居然執行了
|
1
2
|
[www@linux-node1 scripts]$ .
/deploy
.sh rollback
[www@linux-node1 scripts]$
|
改造下
|
1
2
3
4
5
6
7
|
rollback_fun(){
for
node
in
$ROLLBACK_LIST;
do
if
[ -d
/opt/webroot/
$1 ];
then
ssh
$node
"if [ -d /opt/webroot/$1 ];then rm -f /webroot/web-demo && ln -s /opt/webroo/$1 /webroot/web-demo;else echo "
hehe
";fi"
fi
done
}
|
執行
|
1
2
|
[www@linux-node1 scripts]$ .
/deploy
.sh rollback
[www@linux-node1 scripts]$
|
上面執行為和之前顯示一致，再改造下。
|
1
2
3
4
5
|
rollback_fun(){
for
node
in
$ROLLBACK_LIST;
do
ssh
$node
"if [ -d /opt/webroot/$1 ];then rm -f /webroot/web-demo && ln -s /opt/webroo/$1 /webroot/web-demo;else echo "
hehe
";fi"
done
}
|
測試
|
1
2
|
[www@linux-node1 scripts]$ .
/deploy
.sh rollback
[www@linux-node1 scripts]$
|
執行還是沒達到效果
先妥協了。以下面為准。雖然如果少參數，會導致執行不成功，先這樣
|
1
2
3
4
|
rollback_fun(){
for
node
in
$ROLLBACK_LIST;
do
ssh
$node
"rm -f /webroot/web-demo && ln -s /opt/webroo/$1 /webroot/web-demo"
done
}
|
注意把腳本最后一行改成main $1 $2
|
1
2
3
4
5
6
7
8
9
10
11
12
|
[root@linux-node1 scripts]
# tail -10 deploy.sh
shell_lock;
rollback ${ROLLBACK_VER};
shell_unlock;
;;
*)
usage;
esac
}
main $1 $2
[root@linux-node1 scripts]
#
|
繼續測試，修改首頁內容
|
1
2
3
4
5
6
7
8
9
|
[www@linux-node1 web-demo]$
cd
/deploy/code/web-demo
[www@linux-node1 web-demo]$
ls
index.html
[www@linux-node1 web-demo]$
cat
index.html
www.nmap.com
[www@linux-node1 web-demo]$
echo
'www.nmap.org'
>index.html
[www@linux-node1 web-demo]$
cat
index.html
www.nmap.org
[www@linux-node1 web-demo]$
|
繼續執行腳本部署
|
1
2
3
4
5
6
7
8
9
10
11
12
|
[www@linux-node1 web-demo]$
cd
/scripts/
[www@linux-node1 scripts]$ .
/deploy
.sh deploy
git pull
code_build
web-demo_456_2017-04-28-21-17-17.
tar
.gz 100% 222 0.2KB
/s
00:00
web-demo_456_2017-04-28-21-17-17.
tar
.gz 100% 222 0.2KB
/s
00:00
HTTP
/1
.1 200 OK
add to cluster
192.168.58.12.
crontab
.xml 100% 21 0.0KB
/s
00:00
HTTP
/1
.1 200 OK
add to cluster
[www@linux-node1 scripts]$
|
檢查頁面
列出回滾版本
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
|
[www@linux-node1 scripts]$ .
/deploy
.sh rollback list
-rw-rw-r-- 1 www www 204 Apr 23 23:33
/opt/webroot/web-demo_123_2017-04-23-23-33-50
.
tar
.gz
-rw-rw-r-- 1 www www 204 Apr 23 23:43
/opt/webroot/web-demo_123_2017-04-23-23-43-48
.
tar
.gz
-rw-rw-r-- 1 www www 204 Apr 24 00:00
/opt/webroot/web-demo_123_2017-04-24-00-00-14
.
tar
.gz
-rw-rw-r-- 1 www www 204 Apr 24 00:01
/opt/webroot/web-demo_123_2017-04-24-00-01-24
.
tar
.gz
-rw-rw-r-- 1 www www 205 Apr 24 00:02
/opt/webroot/web-demo_123_2017-04-24-00-02-44
.
tar
.gz
-rw-rw-r-- 1 www www 204 Apr 24 00:04
/opt/webroot/web-demo_456_2017-04-24-00-04-05
.
tar
.gz
-rw-rw-r-- 1 www www 220 Apr 27 21:40
/opt/webroot/web-demo_456_2017-04-27-21-40-36
.
tar
.gz
-rw-rw-r-- 1 www www 214 Apr 27 22:41
/opt/webroot/web-demo_456_2017-04-27-22-41-54
.
tar
.gz
-rw-rw-r-- 1 www www 220 Apr 27 23:17
/opt/webroot/web-demo_456_2017-04-27-23-17-27
.
tar
.gz
-rw-rw-r-- 1 www www 220 Apr 28 20:15
/opt/webroot/web-demo_456_2017-04-28-20-15-56
.
tar
.gz
-rw-rw-r-- 1 www www 220 Apr 28 20:16
/opt/webroot/web-demo_456_2017-04-28-20-16-06
.
tar
.gz
-rw-rw-r-- 1 www www 221 Apr 28 20:24
/opt/webroot/web-demo_456_2017-04-28-20-24-23
.
tar
.gz
-rw-rw-r-- 1 www www 220 Apr 28 20:25
/opt/webroot/web-demo_456_2017-04-28-20-25-47
.
tar
.gz
-rw-rw-r-- 1 www www 220 Apr 28 20:28
/opt/webroot/web-demo_456_2017-04-28-20-28-03
.
tar
.gz
-rw-rw-r-- 1 www www 221 Apr 28 20:35
/opt/webroot/web-demo_456_2017-04-28-20-35-37
.
tar
.gz
-rw-rw-r-- 1 www www 222 Apr 28 20:43
/opt/webroot/web-demo_456_2017-04-28-20-43-32
.
tar
.gz
-rw-rw-r-- 1 www www 221 Apr 28 20:45
/opt/webroot/web-demo_456_2017-04-28-20-45-23
.
tar
.gz
-rw-rw-r-- 1 www www 220 Apr 28 20:48
/opt/webroot/web-demo_456_2017-04-28-20-48-01
.
tar
.gz
-rw-rw-r-- 1 www www 222 Apr 28 21:14
/opt/webroot/web-demo_456_2017-04-28-21-14-49
.
tar
.gz
-rw-rw-r-- 1 www www 222 Apr 28 21:17
/opt/webroot/web-demo_456_2017-04-28-21-17-17
.
tar
.gz
[www@linux-node1 scripts]$
|
回滾上一個版本web-demo_456_2017-04-28-21-14-49
|
1
2
|
[www@linux-node1 scripts]$ .
/deploy
.sh rollback web-demo_456_2017-04-28-21-14-49
[www@linux-node1 scripts]$
|
測試訪問
回滾成功，此時腳本內容如下
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
|
[root@linux-node1 scripts]
# cat deploy.sh
#!/bin/bash
#Dir List
#mkdir -p /deploy/code/web-demo/
#mkdir -p /deploy/config/web-demo/base
#mkdir -p /deploy/config/web-demo/other
#mkdir -p /deploy/tar
#mkdir -p /deploy/tmp
#mkdir -p /opt/webroot
#mkdir -p /webroot
#chown -R www:www /deploy
#chown -R www:www /opt/webroot
#chown -R www:www /webroot
#Node List
PRE_LIST=
"192.168.58.11"
GROUP1_LIST=
"192.168.58.12"
ROLLBACK_LIST=
"192.168.58.11 192.168.58.12"
# Date/Time Veriables
LOG_DATE=
'date "+%Y-%m-%d"'
LOG_TIME=
'date "+%H-%M-%S"'
CDATE=$(
date
"+%Y-%m-%d"
)
CTIME=$(
date
"+%H-%M-%S"
)
#Shell Env
SHELL_NAME=
"deploy.sh"
SHELL_DIR=
"/home/www"
SHELL_LOG=
"${SHELL_DIR}/${SHELL_NAME}.log"
#Code Env
PRO_NAME=
"web-demo"
CODE_DIR=
"/deploy/code/web-demo"
CONFIG_DIR=
"/deploy/config/web-demo"
TMP_DIR=
"/deploy/tmp"
TAR_DIR=
"/deploy/tar"
LOCK_FILE=
"/tmp/deploy.lock"
usage(){
echo
$
"Usage: $0 { deploy | rollback [ list |version ]}"
}
writelog(){
LOGINFO=$1
echo
"${CDATE} ${CTIME}: ${SHELL_NAME} : ${LOGINFO}"
>> ${SHELL_LOG}
}
shell_lock(){
touch
${LOCK_FILE}
}
url_test(){
URL=$1
curl -s --
head
$URL |
grep
"200 OK"
if
[ $? -
ne
0 ];
then
shell_unlock;
writelog
"test error"
&&
exit
;
fi
}
shell_unlock(){
rm
-f ${LOCK_FILE}
}
code_get(){
writelog
"code_get"
;
cd
$CODE_DIR &&
echo
"git pull"
cp
-r ${CODE_DIR} ${TMP_DIR}/
API_VER=
"456"
}
code_build(){
echo
code_build
}
code_config(){
writelog
"code_config"
/bin/cp
-r ${CONFIG_DIR}
/base/
* ${TMP_DIR}/
"${PRO_NAME}"
PKG_NAME=
"${PRO_NAME}"
_
"${API_VER}"
_
"${CDATE}-${CTIME}"
cd
${TMP_DIR} &&
mv
${PRO_NAME} ${PKG_NAME}
}
code_tar(){
writelog
"code_tar"
cd
${TMP_DIR} &&
tar
cfz ${PKG_NAME}.
tar
.gz ${PKG_NAME}
writelog
"${PKG_NAME}.tar.gz"
}
code_scp(){
writelog
"code_scp"
for
node
in
$PRE_LIST;
do
scp
${TMP_DIR}/${PKG_NAME}.
tar
.gz $node:
/opt/webroot/
done
for
node
in
$GROUP1_LIST;
do
scp
${TMP_DIR}/${PKG_NAME}.
tar
.gz $node:
/opt/webroot/
done
}
pre_deploy(){
writelog
"remove from cluster"
ssh
$PRE_LIST
"cd /opt/webroot/ && tar xfz ${PKG_NAME}.tar.gz"
ssh
$PRE_LIST
"rm -rf /webroot/web-demo && ln -s /opt/webroot/${PKG_NAME} /webroot/web-demo"
}
pre_test(){
url_test
"http://${PRE_LIST}/index.html"
echo
"add to cluster"
}
group1_deploy(){
writelog
"remove from cluster"
for
node
in
$GROUP1_LIST;
do
ssh
$node
"cd /opt/webroot/ && tar xfz ${PKG_NAME}.tar.gz"
ssh
$node
"rm -rf /webroot/web-demo && ln -s /opt/webroot/${PKG_NAME} /webroot/web-demo"
done
scp
${CONFIG_DIR}
/other/192
.168.58.12.
crontab
.xml 192.168.58.12:
/webroot/web-demo/crontab
.xml
}
group1_test(){
url_test
"http://192.168.58.12/index.html"
echo
"add to cluster"
}
rollback_fun(){
for
node
in
$ROLLBACK_LIST;
do
ssh
$node
"rm -f /webroot/web-demo && ln -s /opt/webroot/$1 /webroot/web-demo"
done
}
rollback(){
if
[ -z $1 ];
then
shell_unlock;
echo
"Please input rollback version"
&&
exit
;
fi
case
$1
in
list)
ls
-l
/opt/webroot/
*.
tar
.gz
;;
*)
rollback_fun $1
esac
}
main(){
if
[ -f ${LOCK_FILE} ];
then
echo
"Deploy is running"
&&
exit
;
fi
DEPLOY_METHOD=$1
ROLLBACK_VER=$2
case
$DEPLOY_METHOD
in
deploy)
shell_lock;
code_get;
code_build;
code_config;
code_tar;
code_scp;
pre_deploy;
pre_test;
group1_deploy;
group1_test;
shell_unlock;
;;
rollback)
shell_lock;
rollback ${ROLLBACK_VER};
shell_unlock;
;;
*)
usage;
esac
}
main $1 $2
[root@linux-node1 scripts]
#
|
再完善下腳本，rollback的時候，后面參數為空，直接退出。也就是整個腳本的$2 為空，退出腳本
|
1
2
3
4
5
6
7
8
9
10
11
12
13
|
rollback(){
if
[ -z $1 ];
then
shell_unlock;
echo
"Please input rollback version"
&&
exit
;
fi
case
$1
in
list)
ls
-l
/opt/webroot/
*.
tar
.gz
;;
*)
rollback_fun $1
esac
}
|
自動化部署實戰-gitlab和部署和回滾
把gitllab裝上，參照鏈接
GitLab是一個利用 Ruby on Rails 開發的開源應用程序，實現一個自托管的Git項目倉庫，可通過Web界面進行訪問公開的或者私人項目。
GitLab擁有與Github類似的功能，能夠瀏覽源代碼，管理缺陷和注釋。可以管理團隊對倉庫的訪問，它非常易於瀏覽提交過的版本並提供一個文件歷史庫。
它還提供一個代碼片段收集功能可以輕松實現代碼復用，便於日后有需要的時候進行查找。
1.基礎環境准備
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
|
[root@linux-node1 ~]
# yum install curl policycoreutils openssh-server openssh-clients postfix -y
Loaded plugins: fastestmirror
Loading mirror speeds from cached hostfile
* base: mirrors.163.com
* epel: mirror01.idc.hinet.net
* extras: mirrors.aliyun.com
* updates: mirrors.163.com
Package curl-7.29.0-35.el7.centos.x86_64 already installed and latest version
Package policycoreutils-2.5-11.el7_3.x86_64 already installed and latest version
Package openssh-server-6.6.1p1-35.el7_3.x86_64 already installed and latest version
Package openssh-clients-6.6.1p1-35.el7_3.x86_64 already installed and latest version
Package 2:postfix-2.10.1-6.el7.x86_64 already installed and latest version
Nothing to
do
[root@linux-node1 ~]
#
|
啟動postfix
|
1
2
|
[root@linux-node1 ~]
# systemctl start postfix
[root@linux-node1 ~]
#
|
2.安裝gitlab-ce
注：由於網絡問題，國內用戶，建議使用163或者阿里雲鏡像源進行安裝：
3.配置並啟動gitlab-ce
默認從163的源里下載，清華大學的鏡像源makecache時報錯
此包286MB，安裝時占800MB多。生產中注意下
|
1
2
3
4
5
6
7
8
9
10
|
[root@linux-node1 ~]
# yum install gitlab-ce -y
Loaded plugins: fastestmirror
Loading mirror speeds from cached hostfile
* base: mirrors.163.com
* epel: mirrors.ustc.edu.cn
* extras: mirrors.163.com
* updates: mirrors.163.com
Package gitlab-ce-9.1.0-ce.0.el7.x86_64 already installed and latest version
Nothing to
do
[root@linux-node1 ~]
#
|
4.配置並啟動gitlab-ce
|
1
|
[root@linux-node1 ~]
# gitlab-ctl reconfigure
|
可以使用gitlab-ctl管理gitlab，例如查看gitlab狀態：
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
|
[root@linux-node1 ~]
# gitlab-ctl status
run: gitaly: (pid 46783) 154s; run: log: (pid 46524) 194s
run: gitlab-monitor: (pid 46839) 152s; run: log: (pid 46671) 170s
run: gitlab-workhorse: (pid 46793) 154s; run: log: (pid 46537) 188s
run: logrotate: (pid 46570) 186s; run: log: (pid 46569) 186s
run: nginx: (pid 47163) 2s; run: log: (pid 46544) 187s
run: node-exporter: (pid 46627) 179s; run: log: (pid 46626) 179s
run: postgres-exporter: (pid 46826) 153s; run: log: (pid 46657) 171s
run: postgresql: (pid 46369) 233s; run: log: (pid 46368) 233s
run: prometheus: (pid 46811) 153s; run: log: (pid 46607) 180s
run: redis: (pid 46312) 239s; run: log: (pid 46311) 239s
run: redis-exporter: (pid 46642) 177s; run: log: (pid 46641) 177s
run: sidekiq: (pid 46509) 195s; run: log: (pid 46508) 195s
run: unicorn: (pid 46483) 196s; run: log: (pid 46482) 196s
[root@linux-node1 ~]
#
|
一些常用命令
|
1
2
3
4
|
gitlab-ctl status
gitlab-ctl stop
gitlab-ctl start
gitlab-ctl restart
|
先停止apache，因為gitlab占用80端口。防止沖突
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
|
[root@linux-node1 ~]
# systemctl stop httpd
[root@linux-node1 ~]
# gitlab-ctl restart
ok: run: gitaly: (pid 47781) 1s
ok: run: gitlab-monitor: (pid 47788) 0s
ok: run: gitlab-workhorse: (pid 47791) 1s
ok: run: logrotate: (pid 47808) 0s
ok: run: nginx: (pid 47814) 1s
ok: run: node-exporter: (pid 47823) 0s
ok: run: postgres-exporter: (pid 47829) 0s
ok: run: postgresql: (pid 47847) 0s
ok: run: prometheus: (pid 47855) 1s
ok: run: redis: (pid 47866) 0s
ok: run: redis-exporter: (pid 47870) 1s
ok: run: sidekiq: (pid 47877) 0s
ok: run: unicorn: (pid 47882) 1s
[root@linux-node1 ~]
# netstat -lntp
Active Internet connections (only servers)
Proto Recv-Q Send-Q Local Address Foreign Address State PID
/Program
name
tcp 0 0 127.0.0.1:9121 0.0.0.0:* LISTEN 47870
/redis_exporte
tcp 0 0 127.0.0.1:9090 0.0.0.0:* LISTEN 47855
/prometheus
tcp 0 0 127.0.0.1:9187 0.0.0.0:* LISTEN 47829
/postgres_expo
tcp 0 0 0.0.0.0:873 0.0.0.0:* LISTEN 1896
/rsync
tcp 0 0 127.0.0.1:9100 0.0.0.0:* LISTEN 47823
/node_exporter
tcp 0 0 0.0.0.0:111 0.0.0.0:* LISTEN 1
/systemd
tcp 0 0 0.0.0.0:80 0.0.0.0:* LISTEN 47814
/nginx
: master
tcp 0 0 127.0.0.1:9168 0.0.0.0:* LISTEN 47788
/ruby
tcp 0 0 192.168.122.1:53 0.0.0.0:* LISTEN 1346
/dnsmasq
tcp 0 0 0.0.0.0:22 0.0.0.0:* LISTEN 45266
/sshd
tcp 0 0 127.0.0.1:25 0.0.0.0:* LISTEN 1284
/master
tcp 0 0 0.0.0.0:8060 0.0.0.0:* LISTEN 47814
/nginx
: master
tcp 0 0 127.0.0.1:25151 0.0.0.0:* LISTEN 3249
/python2
tcp6 0 0 :::873 :::* LISTEN 1896
/rsync
tcp6 0 0 :::3306 :::* LISTEN 41106
/mysqld
tcp6 0 0 :::111 :::* LISTEN 1
/systemd
tcp6 0 0 ::1:9168 :::* LISTEN 47788
/ruby
tcp6 0 0 :::22 :::* LISTEN 45266
/sshd
tcp6 0 0 ::1:25 :::* LISTEN 1284
/master
[root@linux-node1 ~]
#
|
###################################################################################
登錄gitlab
第一次登錄gitlab，需要為root用戶修改密碼，root用戶也是gitlab的超級管理員。
http://192.168.58.11
設置密碼是gitlab
密碼不夠長，重新設置下，設置密碼是12345678，生產中注意密碼復雜點
管理gitlab
使用root用戶和剛才創建的密碼登錄后，你就可以探索gitlab的奧秘了，可以點擊圖中紅框的按鈕進入管理區域。
登錄成功界面
gitlab日常管理的話，其實很少，就是每天進行備份
每小時備份也可以，假如更新頻繁的話
去掉注冊的功能，不讓人注冊，正常應該不讓人隨便注冊
拉到下面
找到下面位置，去掉勾，保存
千萬不要找錯地方了。是Sign-up 而不是Sign-in
再次推出登錄，就沒注冊那一項了
創建個項目之前需要先創建一個組
先創建一個組
填寫如下
訪問級別，設置為內部
創建成功
在組里面創建項目
這里就可以從下拉列表選擇了
項目描述
Import project from 這里可以看到它還可以從其它平台導入
可以從github遷移過來
可以通過key連接它
隨便輸入一串
提交
添加成功
把代碼拉下來
通過ssh拉下來
現在上傳和克隆都不行，你需要配置驗證
把公鑰放上去
|
1
2
3
4
5
6
7
8
9
10
11
12
|
[root@linux-node1 ~]
# cd .ssh/
[root@linux-node1 .
ssh
]
# ll
total 12
-rw------- 1 root root 1675 Apr 5 03:38 id_rsa
-rw-r--r-- 1 root root 407 Apr 5 03:38 id_rsa.pub
-rw-r--r-- 1 root root 175 Apr 4 17:23 known_hosts
[root@linux-node1 .
ssh
]
# cat id_rsa.pub
ssh
-rsa
AAAAB3NzaC1yc2EAAAADAQABAAABAQDYSbAotNl0aNwx09agLbM09A1CCBphaE+g85mQMEnunzg3HSxm6wuFWcDCkK8szqTkMuxneEQOLNwQYsKuEz0pVZlTkN+N38SyhSaMrlCTp8
j3691sbsvMF4owO6jdZpevVoRjaU5v2HmS7yY
/NJh7x1F8TTcfgLQt2rwL6ln02lV1glP1crGmpt95bcKBbnwU4o8W0Q
+Ay4trqPbGJA9sZfU2j4Aanp+pksHt
/yFNnQs2Oy4fXb9c
iQlP5Ue62adN92duMrtGhbo+WevJeEVpqyPbptP906toykpsERb+6GaJwWy
/GtMj/b8lwEnAOW1sY5TIkMb34eV3rXX89L3F
root@linux-node1.nmap.com
[root@linux-node1 .
ssh
]
#
|
title隨便寫，這里是自動生成的
添加成功
查看項目
看到沒ssh-key的提示了
復制git@192.168.58.11:web/web-demo.git
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
|
遇到下面問題
[www@linux-node1 code]$ git
-
bash
: git:
command
not found
安裝git即可
[root@linux-node1 scripts]
# yum install git
Loaded plugins: fastestmirror
Loading mirror speeds from cached hostfile
* base: mirrors.163.com
* epel: mirrors.tuna.tsinghua.edu.cn
* extras: mirrors.163.com
* updates: mirrors.zju.edu.cn
Package git-1.8.3.1-6.el7_2.1.x86_64 already installed and latest version
Nothing to
do
[root@linux-node1 scripts]
#
|
由於實驗環境變化，上面的ssh-key是5天前的，環境變化導致key失效，重新配置key
刪除之前的重新添加
拉取代碼（在這之前，先刪除code目錄下的web-demo）
|
1
2
3
4
5
6
7
8
9
10
11
12
13
|
[www@linux-node1 code]$ git clone git@192.168.58.11:web
/web-demo
.git
Cloning into
'web-demo'
...
remote: Counting objects: 3,
done
.
remote: Total 3 (delta 0), reused 0 (delta 0)
Receiving objects: 100% (3
/3
),
done
.
[www@linux-node1 code]$ ll
total 0
drwxrwxr-x 3 www www 33 Apr 28 21:49 web-demo
[www@linux-node1 code]$
cd
web-demo/
[www@linux-node1 web-demo]$ ll
total 4
-rw-rw-r-- 1 www www 9 Apr 28 21:49 README.md
[www@linux-node1 web-demo]$
|
在家目錄下也克隆一份
|
1
2
3
4
5
6
7
8
9
10
|
[www@linux-node1 web-demo]$
cd
[www@linux-node1 ~]$ git clone git@192.168.58.11:web
/web-demo
.git
Cloning into
'web-demo'
...
remote: Counting objects: 3,
done
.
remote: Total 3 (delta 0), reused 0 (delta 0)
Receiving objects: 100% (3
/3
),
done
.
[www@linux-node1 ~]$
cd
web-demo/
[www@linux-node1 web-demo]$
ls
README.md
[www@linux-node1 web-demo]$
|
添加index.html
|
1
2
3
4
5
6
7
8
9
10
11
|
[www@linux-node1 ~]$
cd
web-demo/
[www@linux-node1 web-demo]$
ls
README.md
[www@linux-node1 web-demo]$
echo
'hehehehhehehe'
>index.html
[www@linux-node1 web-demo]$
cat
index.html
hehehehhehehe
[www@linux-node1 web-demo]$ ll
total 8
-rw-rw-r-- 1 www www 14 Apr 28 21:54 index.html
-rw-rw-r-- 1 www www 9 Apr 28 21:53 README.md
[www@linux-node1 web-demo]$
|
提交時提示需要配置郵箱等
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
|
[www@linux-node1 web-demo]$ git add *
[www@linux-node1 web-demo]$ git commit -m
"add index.html"
*** Please tell me
who
you are.
Run
git config --global user.email
"you@example.com"
git config --global user.name
"Your Name"
to
set
your account's default identity.
Omit --global to
set
the identity only
in
this repository.
fatal: empty ident name (
for
<www@linux-node1.nmap.com>) not allowed
[www@linux-node1 web-demo]$
|
隨便寫個就行
|
1
2
3
4
5
6
|
[www@linux-node1 web-demo]$ git config --global user.email
"nmap@test.com"
[www@linux-node1 web-demo]$ git config --global user.name
"nmap"
[www@linux-node1 web-demo]$ git commit -m
"add index.html"
[master 8c8b197] add index.html
1
file
changed, 1 insertion(+)
create mode 100644 index.html
|
push到gitlab上
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
|
[www@linux-node1 web-demo]$ git push
warning: push.default is
unset
; its implicit value is changing
in
Git 2.0 from
'matching'
to
'simple'
. To squelch this message
and maintain the current behavior after the default changes, use:
git config --global push.default matching
To squelch this message and adopt the new behavior now, use:
git config --global push.default simple
See
'git help config'
and search
for
'push.default'
for
further information.
(the
'simple'
mode was introduced
in
Git 1.7.11. Use the similar mode
'current'
instead of
'simple'
if
you sometimes use older versions of Git)
Counting objects: 4,
done
.
Delta compression using up to 4 threads.
Compressing objects: 100% (2
/2
),
done
.
Writing objects: 100% (3
/3
), 278 bytes | 0 bytes
/s
,
done
.
Total 3 (delta 0), reused 0 (delta 0)
To git@192.168.58.11:web
/web-demo
.git
a2c3614..8c8b197 master -> master
[www@linux-node1 web-demo]$
|
在code目錄拉一下。git pull
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
|
[www@linux-node1 ~]$
cd
/deploy/code/
[www@linux-node1 code]$
cd
web-demo/
[www@linux-node1 web-demo]$ ll
total 4
-rw-rw-r-- 1 www www 9 Apr 28 21:49 README.md
[www@linux-node1 web-demo]$
pwd
/deploy/code/web-demo
[www@linux-node1 web-demo]$ git pull
remote: Counting objects: 3,
done
.
remote: Compressing objects: 100% (2
/2
),
done
.
remote: Total 3 (delta 0), reused 0 (delta 0)
Unpacking objects: 100% (3
/3
),
done
.
From 192.168.58.11:web
/web-demo
a2c3614..8c8b197 master -> origin
/master
Updating a2c3614..8c8b197
Fast-forward
index.html | 1 +
1
file
changed, 1 insertion(+)
create mode 100644 index.html
[www@linux-node1 web-demo]$
[www@linux-node1 web-demo]$ ll
total 8
-rw-rw-r-- 1 www www 14 Apr 28 21:57 index.html
-rw-rw-r-- 1 www www 9 Apr 28 21:49 README.md
[www@linux-node1 web-demo]$
cat
index.html
hehehehhehehe
[www@linux-node1 web-demo]$
|
查看版本號
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
|
[www@linux-node1 web-demo]$ git show
commit 8c8b1977f52f30c55d670f53b3be735a189c6f75
Author: nmap <nmap@
test
.com>
Date: Fri Apr 28 21:56:02 2017 +0800
add index.html
diff
--git a
/index
.html b
/index
.html
new
file
mode 100644
index 0000000..35811e7
---
/dev/null
+++ b
/index
.html
@@ -0,0 +1 @@
+hehehehhehehe
[www@linux-node1 web-demo]$
|
獲取版本號
|
1
2
3
4
5
6
|
[www@linux-node1 web-demo]$ git show |
grep
commit
commit 8c8b1977f52f30c55d670f53b3be735a189c6f75
[www@linux-node1 web-demo]$
[www@linux-node1 web-demo]$ git show |
grep
commit |
cut
-d
' '
-f2
8c8b1977f52f30c55d670f53b3be735a189c6f75
[www@linux-node1 web-demo]$
|
這個git id很長，我們一般取前5、6、8位。很少見前5位重復的。字符串截取
|
1
2
3
4
5
6
|
[www@linux-node1 web-demo]$ git show |
grep
commit |
cut
-d
' '
-f2
8c8b1977f52f30c55d670f53b3be735a189c6f75
[www@linux-node1 web-demo]$ API_VERL=$(git show |
grep
commit |
cut
-d
' '
-f2)
[www@linux-node1 web-demo]$
echo
${API_VERL:0:6}
8c8b19
[www@linux-node1 web-demo]$
|
修改之前代碼，這里直接執行git pull。不執行echo了
|
1
2
3
4
5
6
|
code_get(){
writelog
"code_get"
;
cd
$CODE_DIR && git pull
cp
-r ${CODE_DIR} ${TMP_DIR}/
API_VER=
"456"
}
|
版本也自動獲取
|
1
2
3
4
5
6
7
|
code_get(){
writelog
"code_get"
;
cd
$CODE_DIR && git pull
cp
-r ${CODE_DIR} ${TMP_DIR}/
API_VERL=$(git show |
grep
commit |
cut
-d
' '
-f2)
API_VER=$(
echo
${API_VERL:0:6})
}
|
保存，退出。重新部署下。看到如下，版本變了規整了
|
1
2
3
4
5
6
7
8
9
|
[www@linux-node1 web-demo]$
cd
/scripts/
[www@linux-node1 scripts]$ .
/deploy
.sh
Usage: .
/deploy
.sh { deploy | rollback [ list |version ]}
[www@linux-node1 scripts]$ .
/deploy
.sh deploy
Already up-to-
date
.
code_build
web-demo_8c8b19_2017-04-28-22-03-56.
tar
.gz 100% 8015 7.8KB
/s
00:00
web-demo_8c8b19_2017-04-28-22-03-56.
tar
.gz 100% 8015 7.8KB
/s
00:00
[www@linux-node1 scripts]$
|
服務器都有這個包，沒解壓，網頁訪問沒變化。
檢查下腳本
node1解壓了。但是node2沒解壓
|
1
2
3
4
5
6
7
8
9
10
11
12
|
drwxr-xr-x 2 www www 40 Apr 28 21:17 web-demo_456_2017-04-28-21-17-17
-rw-rw-r-- 1 www www 222 Apr 28 21:17 web-demo_456_2017-04-28-21-17-17.
tar
.gz
drwxrwxr-x 3 www www 67 Apr 28 22:03 web-demo_8c8b19_2017-04-28-22-03-56
-rw-rw-r-- 1 www www 8015 Apr 28 22:03 web-demo_8c8b19_2017-04-28-22-03-56.
tar
.gz
[www@linux-node1 scripts]$ ll
/opt/webroot/
[www@linux-node2 ~]$ ll
/opt/webroot/
-rw-rw-r-- 1 www www 222 Apr 28 21:14 web-demo_456_2017-04-28-21-14-49.
tar
.gz
drwxr-xr-x 2 www www 58 Apr 28 21:17 web-demo_456_2017-04-28-21-17-17
-rw-rw-r-- 1 www www 222 Apr 28 21:17 web-demo_456_2017-04-28-21-17-17.
tar
.gz
-rw-rw-r-- 1 www www 8015 Apr 28 22:03 web-demo_8c8b19_2017-04-28-22-03-56.
tar
.gz
[www@linux-node2 ~]$
|
找到原因了。因為啟動了git。它占用80端口。
而我們先部署的node1，而node1是先測試，通過才能繼續后面的解壓部署操作
node1肯定測試沒通過。因為80端口給git了。
node2理所當然沒解壓了
腳本沒問題
url_test里面加個提示。測試不通過顯示在控制台
|
1
2
3
4
5
6
7
8
|
url_test(){
URL=$1
curl -s --
head
$URL |
grep
"200 OK"
if
[ $? -
ne
0 ];
then
shell_unlock;
writelog
"test error"
&&
exit
;
fi
}
|
改成這個即可
|
1
2
3
4
5
6
7
8
|
url_test(){
URL=$1
curl -s --
head
$URL |
grep
"200 OK"
if
[ $? -
ne
0 ];
then
shell_unlock;
echo
"test error"
&&
exit
;
fi
}
|
再次部署，就提示error了
|
1
2
3
4
5
6
7
|
[www@linux-node1 scripts]$ .
/deploy
.sh deploy
Already up-to-
date
.
code_build
web-demo_8c8b19_2017-04-28-22-07-08.
tar
.gz 100% 8015 7.8KB
/s
00:00
web-demo_8c8b19_2017-04-28-22-07-08.
tar
.gz 100% 8015 7.8KB
/s
00:00
test
error
[www@linux-node1 scripts]$
|
git在用80端口
這里測試的話，可以在main函數里把node1先去掉
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
|
[root@linux-node1 scripts]
# tail -31 deploy.sh
main(){
if
[ -f ${LOCK_FILE} ];
then
echo
"Deploy is running"
&&
exit
;
fi
DEPLOY_METHOD=$1
ROLLBACK_VER=$2
case
$DEPLOY_METHOD
in
deploy)
shell_lock;
code_get;
code_build;
code_config;
code_tar;
code_scp;
# pre_deploy;
# pre_test;
group1_deploy;
group1_test;
shell_unlock;
;;
rollback)
shell_lock;
rollback ${ROLLBACK_VER};
shell_unlock;
;;
*)
usage;
esac
}
main $1 $2
[root@linux-node1 scripts]
#
|
再次執行部署腳本
|
1
2
3
4
5
6
7
8
9
|
[www@linux-node1 scripts]$ .
/deploy
.sh deploy
Already up-to-
date
.
code_build
web-demo_8c8b19_2017-04-28-22-09-43.
tar
.gz 100% 8018 7.8KB
/s
00:00
web-demo_8c8b19_2017-04-28-22-09-43.
tar
.gz 100% 8018 7.8KB
/s
00:00
192.168.58.12.
crontab
.xml 100% 21 0.0KB
/s
00:00
HTTP
/1
.1 200 OK
add to cluster
[www@linux-node1 scripts]$
|
測試OK
修改下主頁文件，再次提交
|
1
2
3
4
5
6
7
8
9
10
11
12
|
[www@linux-node1 ~]$
cd
web-demo/
[www@linux-node1 web-demo]$
ls
index.html README.md
[www@linux-node1 web-demo]$ vim index.html
[www@linux-node1 web-demo]$
cat
index.html
hehehehhehehe
version 2.0
[www@linux-node1 web-demo]$ git add *
[www@linux-node1 web-demo]$ git commit -m
"2.0"
[master 0ac9311] 2.0
1
file
changed, 1 insertion(+)
[www@linux-node1 web-demo]$
|
push上去
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
|
[www@linux-node1 web-demo]$ git push
warning: push.default is
unset
; its implicit value is changing
in
Git 2.0 from
'matching'
to
'simple'
. To squelch this message
and maintain the current behavior after the default changes, use:
git config --global push.default matching
To squelch this message and adopt the new behavior now, use:
git config --global push.default simple
See
'git help config'
and search
for
'push.default'
for
further information.
(the
'simple'
mode was introduced
in
Git 1.7.11. Use the similar mode
'current'
instead of
'simple'
if
you sometimes use older versions of Git)
Counting objects: 5,
done
.
Delta compression using up to 4 threads.
Compressing objects: 100% (2
/2
),
done
.
Writing objects: 100% (3
/3
), 282 bytes | 0 bytes
/s
,
done
.
Total 3 (delta 0), reused 0 (delta 0)
To git@192.168.58.11:web
/web-demo
.git
8c8b197..0ac9311 master -> master
[www@linux-node1 web-demo]$
|
再次執行deploy
它直接從gitlab上拉取代碼了
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
|
[www@linux-node1 scripts]$ .
/deploy
.sh deploy
remote: Counting objects: 3,
done
.
remote: Compressing objects: 100% (2
/2
),
done
.
remote: Total 3 (delta 0), reused 0 (delta 0)
Unpacking objects: 100% (3
/3
),
done
.
From 192.168.58.11:web
/web-demo
76f1c4f..7983661 master -> origin
/master
Updating 76f1c4f..7983661
Fast-forward
index.html | 2 +-
1
file
changed, 1 insertion(+), 1 deletion(-)
code_build
web-demo_798366_2017-04-28-22-21-31.
tar
.gz 100% 9068 8.9KB
/s
00:00
web-demo_798366_2017-04-28-22-21-31.
tar
.gz 100% 9068 8.9KB
/s
00:00
192.168.58.12.
crontab
.xml 100% 21 0.0KB
/s
00:00
HTTP
/1
.1 200 OK
add to cluster
[www@linux-node1 scripts]$
|
成功
回滾到上一個版本
|
1
2
3
4
5
6
7
8
9
|
-rw-rw-r-- 1 www www 222 Apr 28 21:14
/opt/webroot/web-demo_456_2017-04-28-21-14-49
.
tar
.gz
-rw-rw-r-- 1 www www 222 Apr 28 21:17
/opt/webroot/web-demo_456_2017-04-28-21-17-17
.
tar
.gz
-rw-rw-r-- 1 www www 8834 Apr 28 22:19
/opt/webroot/web-demo_76f1c4_2017-04-28-22-19-52
.
tar
.gz
-rw-rw-r-- 1 www www 9068 Apr 28 22:21
/opt/webroot/web-demo_798366_2017-04-28-22-21-31
.
tar
.gz
-rw-rw-r-- 1 www www 8015 Apr 28 22:03
/opt/webroot/web-demo_8c8b19_2017-04-28-22-03-56
.
tar
.gz
-rw-rw-r-- 1 www www 8015 Apr 28 22:07
/opt/webroot/web-demo_8c8b19_2017-04-28-22-07-08
.
tar
.gz
-rw-rw-r-- 1 www www 8018 Apr 28 22:09
/opt/webroot/web-demo_8c8b19_2017-04-28-22-09-43
.
tar
.gz
[www@linux-node1 scripts]$ .
/deploy
.sh rollback web-demo_76f1c4_2017-04-28-22-19-52
[www@linux-node1 scripts]$
|
回滾成功
gitlab相關了解
安裝的配置和目錄結構
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
|
[root@linux-node1 scripts]
# cd /opt/
[root@linux-node1 opt]
# ll
total 12
drwx------ 6 root root 4096 Apr 5 01:11 2017-04-05_01-11-16
drwxr-xr-x 3 root root 32 Apr 5 02:00 full
drwxr-xr-x 10 root root 4096 Apr 5 06:44 gitlab
drwxr-xr-x 4 root root 58 Apr 5 02:03 incr
drwxr-xr-x 23 www www 4096 Apr 28 22:21 webroot
[root@linux-node1 opt]
# cd gitlab/
[root@linux-node1 gitlab]
# ll
total 1944
drwxr-xr-x 2 root root 103 Apr 5 06:43 bin
drwxr-xr-x 18 root root 4096 Apr 5 07:43 embedded
drwxr-xr-x 6 root root 4096 Apr 5 07:43 etc
drwxr-xr-x 2 root root 4096 Apr 5 06:45 init
-rw-r--r-- 1 root root 1931990 Apr 22 21:36 LICENSE
drwxr-xr-x 2 root root 4096 Apr 5 06:43 LICENSES
drwxr-xr-x 2 root root 4096 Apr 5 06:45 service
drwxr-xr-x 15 root root 4096 Apr 5 06:45 sv
drwxr-xr-x 3 root root 20 Apr 5 06:44 var
-rw-r--r-- 1 root root 19834 Apr 22 21:36 version-manifest.json
-rw-r--r-- 1 root root 8769 Apr 22 21:36 version-manifest.txt
[root@linux-node1 gitlab]
#
|
配置文件在etc里面
|
1
2
3
4
5
6
7
8
9
10
11
12
13
|
[root@linux-node1 gitlab]
# cd etc/
[root@linux-node1 etc]
# pwd
/opt/gitlab/etc
[root@linux-node1 etc]
# ll
total 72
drwxr-xr-x 2 root root 28 Apr 5 07:43 gitaly
-rw-r--r-- 1 root root 50 Apr 5 06:44 gitlab-healthcheck-rc
-rw-r--r-- 1 root root 80 Apr 5 06:44 gitlab-psql-rc
drwx------ 3 git root 38 Apr 5 07:43 gitlab-rails
-rw-r--r-- 1 root root 65425 Apr 22 21:36 gitlab.rb.template
drwx------ 3 git root 16 Apr 5 06:44 gitlab-workhorse
drwxr-xr-x 3 root root 16 Apr 5 06:45 postgres-exporter
[root@linux-node1 etc]
#
|
查看下配置目錄
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
|
[root@linux-node1 etc]
# tree
.
├── gitaly
│ ├── HOME
│ └── PATH
├── gitlab-healthcheck-rc
├── gitlab-psql-rc
├── gitlab-rails
│ ├──
env
│ │ ├── BUNDLE_GEMFILE
│ │ ├── EXECJS_RUNTIME
│ │ ├── HOME
│ │ ├── ICU_DATA
│ │ ├── LD_PRELOAD
│ │ ├── PATH
│ │ ├── PYTHONPATH
│ │ ├── RAILS_ENV
│ │ └── SIDEKIQ_MEMORY_KILLER_MAX_RSS
│ └── gitlab-rails-rc
├── gitlab.rb.template
├── gitlab-workhorse
│ └──
env
│ ├── HOME
│ └── PATH
└── postgres-exporter
└──
env
└── DATA_SOURCE_NAME
7 directories, 18 files
[root@linux-node1 etc]
#
|
找它的配置文件
這里可以把external_url之后域名改成IP（早期我就修改了）
|
1
2
3
4
5
6
7
8
9
|
[root@linux-node1 etc]
# rpm -ql gitlab-ce | less
[root@linux-node1 etc]
# cd /etc/gitlab/
[root@linux-node1 gitlab]
# grep external_url gitlab.rb
##! For more details on configuring external_url see:
external_url
'http://192.168.58.11'
# registry_external_url 'https://registry.gitlab.example.com'
# pages_external_url "http://pages.example.com/"
# mattermost_external_url 'http://mattermost.example.com'
[root@linux-node1 gitlab]
#
|
改完之后，需要重新配置
|
1
|
[root@linux-node1 gitlab]
# gitlab-ctl reconfigure
|
重新登錄
如果重新登錄，發現還沒被改掉，應該是改之前的項目沒變，新建的項目用的都是IP了
最終的腳本內容
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
|
[root@linux-node1 gitlab]
# cd /scripts/
[root@linux-node1 scripts]
# mv deploy.sh deploy_all.sh
[root@linux-node1 scripts]
# vim deploy_all.sh
[root@linux-node1 scripts]
# cat deploy_all.sh
#!/bin/bash
#Dir List
#mkdir -p /deploy/code/web-demo/
#mkdir -p /deploy/config/web-demo/base
#mkdir -p /deploy/config/web-demo/other
#mkdir -p /deploy/tar
#mkdir -p /deploy/tmp
#mkdir -p /opt/webroot
#mkdir -p /webroot
#chown -R www:www /deploy
#chown -R www:www /opt/webroot
#chown -R www:www /webroot
#Node List
PRE_LIST=
"192.168.58.11"
GROUP1_LIST=
"192.168.58.12"
ROLLBACK_LIST=
"192.168.58.11 192.168.58.12"
# Date/Time Veriables
LOG_DATE=
'date "+%Y-%m-%d"'
LOG_TIME=
'date "+%H-%M-%S"'
CDATE=$(
date
"+%Y-%m-%d"
)
CTIME=$(
date
"+%H-%M-%S"
)
#Shell Env
SHELL_NAME=
"deploy.sh"
SHELL_DIR=
"/home/www"
SHELL_LOG=
"${SHELL_DIR}/${SHELL_NAME}.log"
#Code Env
PRO_NAME=
"web-demo"
CODE_DIR=
"/deploy/code/web-demo"
CONFIG_DIR=
"/deploy/config/web-demo"
TMP_DIR=
"/deploy/tmp"
TAR_DIR=
"/deploy/tar"
LOCK_FILE=
"/tmp/deploy.lock"
usage(){
echo
$
"Usage: $0 { deploy | rollback [ list |version ]}"
}
writelog(){
LOGINFO=$1
echo
"${CDATE} ${CTIME}: ${SHELL_NAME} : ${LOGINFO}"
>> ${SHELL_LOG}
}
shell_lock(){
touch
${LOCK_FILE}
}
url_test(){
URL=$1
curl -s --
head
$URL |
grep
"200 OK"
if
[ $? -
ne
0 ];
then
shell_unlock;
echo
"test error"
&&
exit
;
fi
}
shell_unlock(){
rm
-f ${LOCK_FILE}
}
code_get(){
writelog
"code_get"
;
cd
$CODE_DIR && git pull
cp
-r ${CODE_DIR} ${TMP_DIR}/
API_VERL=$(git show |
grep
commit |
cut
-d
' '
-f2)
API_VER=$(
echo
${API_VERL:0:6})
}
code_build(){
echo
code_build
}
code_config(){
writelog
"code_config"
/bin/cp
-r ${CONFIG_DIR}
/base/
* ${TMP_DIR}/
"${PRO_NAME}"
PKG_NAME=
"${PRO_NAME}"
_
"${API_VER}"
_
"${CDATE}-${CTIME}"
cd
${TMP_DIR} &&
mv
${PRO_NAME} ${PKG_NAME}
}
code_tar(){
writelog
"code_tar"
cd
${TMP_DIR} &&
tar
cfz ${PKG_NAME}.
tar
.gz ${PKG_NAME}
writelog
"${PKG_NAME}.tar.gz"
}
code_scp(){
writelog
"code_scp"
for
node
in
$PRE_LIST;
do
scp
${TMP_DIR}/${PKG_NAME}.
tar
.gz $node:
/opt/webroot/
done
for
node
in
$GROUP1_LIST;
do
scp
${TMP_DIR}/${PKG_NAME}.
tar
.gz $node:
/opt/webroot/
done
}
pre_deploy(){
writelog
"remove from cluster"
ssh
$PRE_LIST
"cd /opt/webroot/ && tar xfz ${PKG_NAME}.tar.gz"
ssh
$PRE_LIST
"rm -rf /webroot/web-demo && ln -s /opt/webroot/${PKG_NAME} /webroot/web-demo"
}
pre_test(){
url_test
"http://${PRE_LIST}/index.html"
echo
"add to cluster"
}
group1_deploy(){
writelog
"remove from cluster"
for
node
in
$GROUP1_LIST;
do
ssh
$node
"cd /opt/webroot/ && tar xfz ${PKG_NAME}.tar.gz"
ssh
$node
"rm -rf /webroot/web-demo && ln -s /opt/webroot/${PKG_NAME} /webroot/web-demo"
done
scp
${CONFIG_DIR}
/other/192
.168.58.12.
crontab
.xml 192.168.58.12:
/webroot/web-demo/crontab
.xml
}
group1_test(){
url_test
"http://192.168.58.12/index.html"
echo
"add to cluster"
}
rollback_fun(){
for
node
in
$ROLLBACK_LIST;
do
ssh
$node
"rm -f /webroot/web-demo && ln -s /opt/webroot/$1 /webroot/web-demo"
done
}
rollback(){
if
[ -z $1 ];
then
shell_unlock;
echo
"Please input rollback version"
&&
exit
;
fi
case
$1
in
list)
ls
-l
/opt/webroot/
*.
tar
.gz
;;
*)
rollback_fun $1
esac
}
main(){
if
[ -f ${LOCK_FILE} ];
then
echo
"Deploy is running"
&&
exit
;
fi
DEPLOY_METHOD=$1
ROLLBACK_VER=$2
case
$DEPLOY_METHOD
in
deploy)
shell_lock;
code_get;
code_build;
code_config;
code_tar;
code_scp;
pre_deploy;
pre_test;
group1_deploy;
group1_test;
shell_unlock;
;;
rollback)
shell_lock;
rollback ${ROLLBACK_VER};
shell_unlock;
;;
*)
usage;
esac
}
main $1 $2
[root@linux-node1 scripts]
#
|
疑難解答。如果
http://192.168.58.12/提示下面這種
很大可能8080端口被別的占用了
