Chinaunix首页 | 论坛 | 博客

  • 博客访问: 915838
  • 博文数量: 226
  • 博客积分: 10077
  • 博客等级: 上将
  • 技术积分: 2708
  • 用 户 组: 普通用户
  • 注册时间: 2008-05-25 14:36
文章分类

全部博文(226)

文章存档

2010年(15)

2009年(82)

2008年(129)

我的朋友

分类: LINUX

2010-11-16 15:44:09

1.发生琐定,解除步骤:

 A.确定没有存取.关闭httpd服务器.
 B.成为管理员身份
 C.执行svnadmin recover /path/to/repo命令
 D.重启动subversion.

 //注意:
 A.以管理员身份,非root身份,如果一root身份,要重新chmod,库的权限.

2.升级的步骤:

 A.现有版本svnadmin ,将档案库dump出来
 B.升级到新版的subversion
 C.删除就档案库,再原处以新版本subversion,建立新的空档案库.
 D.再利用新版本svnadmin ,将档案库load到刚刚建立好的档案库.

 实例:
 //导出
 $svnlook youngest oldrepo
 $svnadmin dump oldrepo >; dumpfile
 //还原
 $svnadmin load newrepo < dumpfile

 3.即时备份目录

 $svnadmin hotcopy /path/to/repo /backup/repo

4.渐进式备份实现.

A.建立文件hot-backup.perl.内容如下:

 #!/usr/bin/perl -w
 use strict;

my $repos_path = '/path/to/repos';
 my $dumpfile = '/usr/backup/svn-dumpfile';
 my $last_dumped = '/var/log/svn-last-dumped';
 my $svnlook = '/usr/local/subversion/bin/svnlook';
 my $svnadmin = '/usr/local/subversion/bin/svnadmin';

# Figure out the starting revision. Use 0 if we cannot read the
 # last-dumped file, else use the revision in that file incremented
 # by 1.
 my $new_start = 0;
 if (open LASTDUMPED, $last_dumped)
 {
 my $line = ;;
 if (defined $line and $line =~ /^(\d+)/)
 {
 $new_start = $1 + 1;
 }
 close LASTDUMPED;
 }

# Query the youngest revision in the repos.
 my $youngest = `$svnlook youngest $repos_path`;
 defined $youngest && $youngest =~ /^\d+$/
 or die "$0: 'svnlook youngest $repos_path' cannot get youngest revision.\n";
 chomp $youngest;

# Do the backup.
 system("$svnadmin dump $repos_path --revision $new_startyoungest --incremental >;>; $dumpfile.tmp" == 0
 or die "$0: svnadmin dump to '$dumpfile.tmp' failed.\n";

# Store a new last-dumped revision.
 open LASTDUMPED, ">; $last_dumped.tmp"
 or die "$0: cannot open '$last_dumped.tmp' for writing: $!\n";
 print LASTDUMPED "$youngest\n";
 close LASTDUMPED
 or die "$0: error in closing '$last_dumped.tmp' for writing: $!\n";

# Rename to final locations.
 rename("$dumpfile.tmp", $dumpfile)
 or die "$0: cannot rename '$dumpfile.tmp' to '$dumpfile': $!\n";
 rename("$last_dumped.tmp", $last_dumped)
 or die "$0: cannot rename '$last_dumped.tmp' to '$last_dumped': $!\n";
 # All done!

B.备份.先编辑hot-backup.perl里面正确的路径名.

 #vi hot-backup.perl修改为自己的路径:
 my $repos_path = '/path/to/repos';
 my $dumpfile = '/usr/backup/svn-dumpfile';
 my $last_dumped = '/var/log/svn-last-dumped';
 my $svnlook = '/usr/local/subversion/bin/svnlook';
 my $svnadmin = '/usr/local/subversion/bin/svnadmin';
 #perl hot-backup.perl得到的文件是/usr/backup/svn-dumpfile

c.还原

 1.建立新的仓库
 #svnadmin create /usr/local/repo/newrepo

 2.还原信息
 #svnadmin load /usr/local/repo/newrepo < /usr/backup/svn-dumpfile

 3.注意:以后每天都备份.再执行还原即可,不必新建仓库了.

 4.为了每段时间能验证备份的有效性.最好是建立一个验证仓库,用来还原,看信息是否正常。

d.有提交,就备份.进入仓库

 #mv /hooks/post-commit.tmpl /hooks/post-commit
 #vi /hooks/post-commit
 添加内容在后面
 perl /usr/local/subversion/backup/hot-backup.perl
 每次提交触发一次备份

E.修改一下上面的程序,让能接受参数输入

 #!/usr/bin/perl -w
 use strict;

my $repos_path = $ARGV[0];
 my $dumpfile = $ARGV[1];
 my $last_dumped = $ARGV[2];

my $svnlook = '/usr/local/subversion/bin/svnlook';
 my $svnadmin = '/usr/local/subversion/bin/svnadmin';

#for usage
 if(@ARGV != 3 )
 {
 print "$0:[/path/to/repo] [/backup/dumpfile] [/var/log/lastdump]\n";
 exit 1 ;
 }

# Figure out the starting revision. Use 0 if we cannot read the
 # last-dumped file, else use the revision in that file incremented
 # by 1.
 my $new_start = 0;
 if (open LASTDUMPED, $last_dumped)
 {
 my $line = ;;
 if (defined $line and $line =~ /^(\d+)/)
 {
  $new_start = $1 + 1;
 }
 close LASTDUMPED;
 }

# Query the youngest revision in the repos.
 my $youngest = `$svnlook youngest $repos_path`;
 defined $youngest && $youngest =~ /^\d+$/
 or die "$0: 'svnlook youngest $repos_path' cannot get youngest revision.\n";
 chomp $youngest;

# Do the backup.
 system("$svnadmin dump $repos_path --revision $new_startyoungest --incremental >;>; $dumpfile.tmp" == 0
 or die "$0: svnadmin dump to '$dumpfile.tmp' failed.\n";

# Store a new last-dumped revision.
 open LASTDUMPED, ">; $last_dumped.tmp"
 or die "$0: cannot open '$last_dumped.tmp' for writing: $!\n";
 print LASTDUMPED "$youngest\n";
 close LASTDUMPED
 or die "$0: error in closing '$last_dumped.tmp' for writing: $!\n";

# Rename to final locations.
 rename("$dumpfile.tmp", $dumpfile)
 or die "$0: cannot rename '$dumpfile.tmp' to '$dumpfile': $!\n";
 rename("$last_dumped.tmp", $last_dumped)
 or die "$0: cannot rename '$last_dumped.tmp' to '$last_dumped': $!\n";
 # All done!

F.有提交,就备份.进入仓库

 #mv /hooks/post-commit.tmpl /hooks/post-commit
 #vi /hooks/post-commit
 添加内容在后面
 perl /usr/local/subversion/backup/hot-backup.pl /usr/local/repo/webaddress  /usr/local/subversion/backup/webaddress /var/log/webaddress
 每次提交触发一次备份

5.hot-backup.py完整备份

 6.tar包完整备份

 1.进入仓库的根目录/usrl/local/
 #cd /usr/local/

 2.备份整个目录.
 #tar -cf repo.tar /usr/local/repo.

 3.还原时候来到根目录/usrl/local/
 #tar -xvf repo.tar

4.也可以备份单个目录.
 #tar -cf svn.tar /usr/local/repo/svn.

 5.还原单个目录到目录/usr/local/repo/
 #tar -xvf svn.tar

 6.每天的自动备份脚本backup.sh.放于/usrl/local/subversion/backup目录下
 #!/bin/sh
 USAGE="Usage:`basename $0` [-f|-d] [files|directories]"
 if [ $# -lt 2 ] ; then
 echo "$USAGE";
 exit 1 ;
 fi
 case "$1" in
 -f) shift
 TARGES="-tvf" ;
 for i in "$@" ;
 do
 if [ -f "$i" ] ; then
 FILES = `tar $TAGES "$i" 2 >; /dev/null`
 if [ $? -eq 0 ] ; then
 echo ; echo "$i" ; echo "$FILES"
 else
 echo "ERRORi not a tar file."
 fi
 else
 echo "ERRORi not a file.";
 fi
 done
 ;;
 -d) shift
 TARGES="-cvf" ;
 if [ -f backup.tar ] ; then
 mv backup.tar backup1.tar
 fi
 tar $TARGES backup.tar "$@"
 ;;
 *) echo "$USAGE"
 exit 0
 ;;
 esac
 exit $?

8.固定时间运行脚本

 A.建立运行脚本backup:
 20 18 * * * /usr/local/subversion/backup/backup.sh -d /usr/local/repo
 B.启用脚本
 #crontab backup
 #crontab -l 查看运行的脚本
 20 18 * * * /usr/local/subversion/backup/backup.sh -d /usr/local/repo //每日18:20,下班后20分钟

注意:

 1.linux直接cp的备份仓库能在nt下使用,但是nt的复制不能在linux下使用.

阅读(2344) | 评论(1) | 转发(0) |
给主人留下些什么吧!~~

chinaunix网友2010-11-19 15:19:33

很好的, 收藏了 推荐一个博客,提供很多免费软件编程电子书下载: http://free-ebooks.appspot.com