转载：Gitlab备份和恢复操作记录

转载：Gitlab备份和恢复操作记录

　　包含了备份和数据恢复的操作记录，实验可行

前面已经介绍了Gitlab环境部署记录，这里简单说下Gitlab的备份和恢复操作记录：

1）Gitlab的备份目录路径设置

1 2 3 4 5 6 7 8 9 10 11 12

[root@code-server ~]# vim /etc/gitlab/gitlab.rb gitlab_rails['manage_backup_path'] = true gitlab_rails['backup_path'] = "/data/gitlab/backups"    //gitlab备份目录 gitlab_rails['backup_archive_permissions'] = 0644       //生成的备份文件权限 gitlab_rails['backup_keep_time'] = 7776000              //备份保留天数为3个月（即90天，这里是7776000秒）   [root@code-server ~]# mkdir -p /data/gitlab/backups [root@code-server ~]# chown -R git.git /data/gitlab/backups [root@code-server ~]# chmod -R 777 /data/gitlab/backups    如上设置了gitlab备份目录路径为/data/gitlab/backups，最后使用下面命令重载gitlab配置文件，是上述修改生效！ root@code-server ~]# gitlab-ctl reconfigure

2）GItlab备份操作（使用备份命令"gitlab-rake gitlab:backup:create"）

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

手动备份gitlab [root@code-server backups]# gitlab-rake gitlab:backup:create Dumping database ... Dumping PostgreSQL database gitlabhq_production ... [DONE] done Dumping repositories ...  * treesign/treesign ... [DONE]  * gateway/gateway ... [DONE]  * treesign/treesign-doc ... [SKIPPED]  * qwsign/qwsign ... [DONE]  * qwsign/qwsign-doc ... [DONE]  * test/test ... [DONE] done Dumping uploads ... done Dumping builds ... done Dumping artifacts ... done Dumping pages ... done Dumping lfs objects ... done Dumping container registry images ... [DISABLED] Creating backup archive: 1510471890_2017_11_12_9.4.5_gitlab_backup.tar ... done Uploading backup archive to remote storage  ... skipped Deleting tmp directories ... done done done done done done done done Deleting old backups ... done. (0 removed)   然后查看下备份文件（文件权限是设定好的644） [root@code-server backups]# ll total 244 -rw-r--r-- 1 git git 245760 Nov 12 15:33 1510472027_2017_11_12_9.4.5_gitlab_backup.tar   编写备份脚本，结合crontab实施自动定时备份，比如每天0点、6点、12点、18点各备份一次 [root@code-server backups]# pwd /data/gitlab/backups [root@code-server backups]# vim gitlab_backup.sh #!/bin/bash /usr/bin/gitlab-rake gitlab:backup:create CRON=1   注意：环境变量CRON=1的作用是如果没有任何错误发生时， 抑制备份脚本的所有进度输出   [root@code-server backups]# crontab -l 0 0,6,12,18 * * * /bin/bash -x /data/gitlab/backups/gitlab_backup.sh > /dev/null 2>&1

3）Gitlab自动备份

1）.周期性计划任务

　　添加备份信息

$crontab -e
0 0 * * * /opt/gitlab/bin/gitlab-rake gitlab:backup:create

　  重启备份服务　

$service crond restart

　  每天凌晨备份gitlab的数据

2）.删除过期的备份文件
因为每天都会进行一次备份，而备份的数据比较大，磁盘空间会被大量使用，因此，定期删除过期的文件

$vim /var/opt/gitlab/backups/remove.sh

添加下面的内容

!/bin/bash
find "/var/opt/gitlab/backups/" -name ".tar" -ctime +1 -type f -exec rm -rf {} ;

设置每天晚上11点，删除两天前的备份

$contab -e
0 23 * * root /var/opt/gitlab/backups/remove.sh -D 2     #添加内容

赋予脚本执行权限，否则用户不能执行此脚本.

$chmod +x remove.sh

重启备份器

$service crond restart　

4）Gitlab恢复操作

GItlab只能还原到与备份文件相同的gitlab版本。

不同版本的下载可从这里下载：https://packages.gitlab.com/gitlab/

Ubanu16.04选择gitlab-ce系列的安装包，如下图所示，根据原始版本（cat /opt/gitlab/embedded/service/gitlab-rails/VERSION）来下载。

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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189

假设在上面gitlab备份之前创建了test项目，然后不小心误删了test项目，现在就进行gitlab恢复操作：    1）停止相关数据连接服务 [root@code-server backups]# gitlab-ctl stop unicorn ok: down: unicorn: 0s, normally up [root@code-server backups]# gitlab-ctl stop sidekiq ok: down: sidekiq: 1s, normally up [root@code-server backups]# gitlab-ctl status run: gitaly: (pid 98087) 1883s; run: log: (pid 194202) 163003s run: gitlab-monitor: (pid 98101) 1883s; run: log: (pid 194363) 163002s run: gitlab-workhorse: (pid 98104) 1882s; run: log: (pid 194362) 163002s run: logrotate: (pid 98117) 1882s; run: log: (pid 5793) 160832s run: nginx: (pid 98123) 1881s; run: log: (pid 194359) 163002s run: node-exporter: (pid 98167) 1881s; run: log: (pid 194360) 163002s run: postgres-exporter: (pid 98173) 1881s; run: log: (pid 194204) 163003s run: postgresql: (pid 98179) 1880s; run: log: (pid 194365) 163002s run: prometheus: (pid 98187) 1880s; run: log: (pid 194364) 163002s run: redis: (pid 98230) 1879s; run: log: (pid 194358) 163002s run: redis-exporter: (pid 98234) 1879s; run: log: (pid 194208) 163003s down: sidekiq: 8s, normally up; run: log: (pid 194437) 163001s down: unicorn: 21s, normally up; run: log: (pid 194443) 163001s    2）现在通过之前的备份文件进行恢复 [root@code-server backups]# pwd /data/gitlab/backups [root@code-server backups]# ll total 244 -rw-r--r-- 1 git git 245760 Nov 12 15:33 1510472027_2017_11_12_9.4.5_gitlab_backup.tar    Gitlab的恢复操作会先将当前所有的数据清空，然后再根据备份数据进行恢复 [root@code-server backups]# gitlab-rake gitlab:backup:restore BACKUP=1510472027_2017_11_12_9.4.5 Unpacking backup ... done Before restoring the database we recommend removing all existing tables to avoid future upgrade problems. Be aware that if you have custom tables in the GitLab database these tables and all data will be removed.    Do you want to continue (yes/no)? ........ ALTER TABLE ALTER TABLE ALTER TABLE ALTER TABLE WARNING:  no privileges were granted for "public" GRANT [DONE] done Restoring repositories ...  * treesign/treesign ... [DONE]  * gateway/gateway ... [DONE]  * treesign/treesign-doc ... [DONE]  * qwsign/qwsign ... [DONE]  * qwsign/qwsign-doc ... [DONE]  * test/test ... [DONE] Put GitLab hooks in repositories dirs [DONE] done Restoring uploads ... done Restoring builds ... done Restoring artifacts ... done Restoring pages ... done Restoring lfs objects ... done This will rebuild an authorized_keys file. You will lose any data stored in authorized_keys file. Do you want to continue (yes/no)? yes       Deleting tmp directories ... done done done done done done done done [root@code-server backups]#    最后再次启动Gitlab [root@code-server backups]# gitlab-ctl start ok: run: gitaly: (pid 98087) 2138s ok: run: gitlab-monitor: (pid 98101) 2138s ok: run: gitlab-workhorse: (pid 98104) 2137s ok: run: logrotate: (pid 98117) 2137s ok: run: nginx: (pid 98123) 2136s ok: run: node-exporter: (pid 98167) 2136s ok: run: postgres-exporter: (pid 98173) 2136s ok: run: postgresql: (pid 98179) 2135s ok: run: prometheus: (pid 98187) 2135s ok: run: redis: (pid 98230) 2134s ok: run: redis-exporter: (pid 98234) 2134s ok: run: sidekiq: (pid 104494) 0s ok: run: unicorn: (pid 104497) 1s [root@code-server backups]# gitlab-ctl status run: gitaly: (pid 98087) 2142s; run: log: (pid 194202) 163262s run: gitlab-monitor: (pid 98101) 2142s; run: log: (pid 194363) 163261s run: gitlab-workhorse: (pid 98104) 2141s; run: log: (pid 194362) 163261s run: logrotate: (pid 98117) 2141s; run: log: (pid 5793) 161091s run: nginx: (pid 98123) 2140s; run: log: (pid 194359) 163261s run: node-exporter: (pid 98167) 2140s; run: log: (pid 194360) 163261s run: postgres-exporter: (pid 98173) 2140s; run: log: (pid 194204) 163262s run: postgresql: (pid 98179) 2139s; run: log: (pid 194365) 163261s run: prometheus: (pid 98187) 2139s; run: log: (pid 194364) 163261s run: redis: (pid 98230) 2138s; run: log: (pid 194358) 163261s run: redis-exporter: (pid 98234) 2138s; run: log: (pid 194208) 163262s run: sidekiq: (pid 104494) 4s; run: log: (pid 194437) 163260s run: unicorn: (pid 104497) 4s; run: log: (pid 194443) 163260s   恢复命令完成后，可以check检查一下恢复情况 [root@code-server backups]# gitlab-rake gitlab:check SANITIZE=true Checking GitLab Shell ...   GitLab Shell version >= 5.3.1 ? ... OK (5.3.1) Repo base directory exists? default... yes Repo storage directories are symlinks? default... no Repo paths owned by git:root, or git:git? default... yes Repo paths access is drwxrws---? default... yes hooks directories in repos are links: ... 5/1 ... ok 6/2 ... ok 5/3 ... repository is empty 12/4 ... ok 12/5 ... ok Running /opt/gitlab/embedded/service/gitlab-shell/bin/check Check GitLab API access: OK Access to /var/opt/gitlab/.ssh/authorized_keys: OK Send ping to redis server: OK gitlab-shell self-check successful   Checking GitLab Shell ... Finished   Checking Sidekiq ...   Running? ... yes Number of Sidekiq processes ... 1   Checking Sidekiq ... Finished   Checking Reply by email ...   Reply by email is disabled in config/gitlab.yml   Checking Reply by email ... Finished   Checking LDAP ...   LDAP is disabled in config/gitlab.yml   Checking LDAP ... Finished   Checking GitLab ...   Git configured correctly? ... yes Database config exists? ... yes All migrations up? ... yes Database contains orphaned GroupMembers? ... no GitLab config exists? ... yes GitLab config up to date? ... yes Log directory writable? ... yes Tmp directory writable? ... yes Uploads directory exists? ... yes Uploads directory has correct permissions? ... yes Uploads directory tmp has correct permissions? ... yes Init script exists? ... skipped (omnibus-gitlab has no init script) Init script up-to-date? ... skipped (omnibus-gitlab has no init script) Projects have namespace: ... 5/1 ... yes 6/2 ... yes 5/3 ... yes 12/4 ... yes 12/5 ... yes Redis version >= 2.8.0? ... yes Ruby version >= 2.3.3 ? ... yes (2.3.3) Git version >= 2.7.3 ? ... yes (2.13.4) Active users: ... 11   Checking GitLab ... Finished    然后稍等一会（如果启动gitlab后，访问出现500，这是因为redis等程序还没完全启动，等一会儿访问就ok了），再次登录Gitlab，就会发现之前误删除的test项目已经恢复了！    另外：Gitlab迁移与恢复一样，但是要求两个GitLab版本号一致

 

 

 

 

 

 

 

 

赋予脚本执行权限，否则用户不能执行此脚本.

$chmod +x remove.sh