在此处输入代码我第一次尝试使用橡胶将 Rails 3 应用程序部署到 ec2。在命令 cap Rubber:create_staging 运行期间，要求实例重新启动，然后它不会接受连接。这是它的外观：
** [out :: production.foo.com] Setting up grub2-common (1.99-21ubuntu3.1) ...
** [out :: production.foo.com] Setting up grub-pc-bin (1.99-21ubuntu3.1) ...
** [out :: production.foo.com] Setting up grub-pc (1.99-21ubuntu3.1) ...
** [out :: production.foo.com] Generating grub.cfg ...
** [out :: production.foo.com] /usr/sbin/grub-probe: warn:
** [out :: production.foo.com]
** [out :: production.foo.com] disk does not exist, so falling back to partition device /dev/xvda1
** [out :: production.foo.com] .
** [out :: production.foo.com] Found linux image: /boot/vmlinuz-3.2.0-26-virtual
** [out :: production.foo.com] Found initrd image: /boot/initrd.img-3.2.0-26-virtual
** [out :: production.foo.com] /usr/sbin/grub-probe: warn:
** [out :: production.foo.com]
** [out :: production.foo.com] disk does not exist, so falling back to partition device /dev/xvda1
** [out :: production.foo.com] .
** [out :: production.foo.com] /usr/sbin/grub-probe: warn:
** [out :: production.foo.com]
** [out :: production.foo.com] disk does not exist, so falling back to partition device /dev/xvda1.
** [out :: production.foo.com] /usr/sbin/grub-probe: warn:
** [out :: production.foo.com]
** [out :: production.foo.com] disk does not exist, so falling back to partition device /dev/xvda1
** [out :: production.foo.com] .
** [out :: production.foo.com] /usr/sbin/grub-probe: warn:
** [out :: production.foo.com]
** [out :: production.foo.com] disk does not exist, so falling back to partition device /dev/xvda1
** [out :: production.foo.com] .
** [out :: production.foo.com] Found linux image: /boot/vmlinuz-3.2.0-23-virtual
** [out :: production.foo.com] Found initrd image: /boot/initrd.img-3.2.0-23-virtual
** [out :: production.foo.com] /usr/sbin/grub-probe: warn:
** [out :: production.foo.com]
** [out :: production.foo.com] disk does not exist, so falling back to partition device /dev/xvda1
** [out :: production.foo.com] .
** [out :: production.foo.com] /usr/sbin/grub-probe: warn:
** [out :: production.foo.com]
** [out :: production.foo.com] disk does not exist, so falling back to partition device /dev/xvda1
** [out :: production.foo.com] .
** [out :: production.foo.com] /usr/sbin/grub-probe: warn:
** [out :: production.foo.com]
** [out :: production.foo.com] disk does not exist, so falling back to partition device /dev/xvda1
** [out :: production.foo.com] .
** [out :: production.foo.com] /usr/sbin/grub-probe: warn:
** [out :: production.foo.com]
** [out :: production.foo.com] disk does not exist, so falling back to partition device /dev/xvda1
** [out :: production.foo.com] .
** [out :: production.foo.com] /usr/sbin/grub-probe: warn:
** [out :: production.foo.com]
** [out :: production.foo.com] disk does not exist, so falling back to partition device /dev/xvda1
** [out :: production.foo.com] .
** [out :: production.foo.com] Found memtest86+ image: /boot/memtest86+.bin
** [out :: production.foo.com] done
** [out :: production.foo.com] Processing triggers for libc-bin ...
** [out :: production.foo.com] ldconfig deferred processing now taking place
** [out :: production.foo.com] Processing triggers for resolvconf ...
** [out :: production.foo.com] resolvconf: Error: /etc/resolv.conf isn't a symlink, not doing anything.
** [out :: production.foo.com] Processing triggers for initramfs-tools ...
** [out :: production.foo.com] update-initramfs: Generating /boot/initrd.img-3.2.0-26-virtual
command finished in 131854ms
* executing "echo $(ls /var/run/reboot-required 2> /dev/null)"
servers: ["production.foo.com"]
[production.foo.com] executing command
command finished in 460ms
* executing "echo $(ls /mnt/your_app_name-production 2> /dev/null)"
servers: ["production.foo.com"]
[production.foo.com] executing command
command finished in 473ms
** Updates require a reboot on hosts ["production.foo.com"]
** Rebooting ...
* executing "sudo -p 'sudo password: ' reboot"
servers: ["production.foo.com"]
[production.foo.com] executing command
command finished in 479ms
* executing `rubber:_direct_connection_production.foo.com_887'
* executing "echo"
servers: ["production.foo.com"]
** Failed to connect to production.foo.com, retrying
问题是，通过运行 Rubber:create 或任何其他橡胶命令，在创建和初始化实例后，在写入 /etc/hosts 文件后，我收到“连接失败”错误，一切都停在那里。
如果我对写在 hosts 文件中的地址执行 ssh，那么我就可以完美连接到实例，所以我不明白问题出在哪里……</p>