AWS Ubuntu EC2实例误设/home目录777权限后无法连接的恢复方案咨询

Alright, let's get your EC2 instance back online—this is a classic "oops" moment, but AWS gives us a straightforward way to recover without losing your data. Here's what to do:

Step 1: Prepare the Faulty Instance's Root Volume #

• Stop your problematic EC2 instance (don't terminate it—we need the data!).
• Go to the EC2 Console > Volumes, find the root volume attached to your instance (it'll show as "/dev/sda1" or "/dev/xvda" in the "Attachment Information" column).
• Detach this volume from the instance (right-click > Detach Volume).

Step 2: Launch a Temporary Rescue Instance #

• Spin up a new Ubuntu EC2 instance in the same Availability Zone as your original instance (this is critical for attaching the volume later).
• Use the same Ubuntu version as your original instance if possible—avoids permission/tool mismatches.

Step 3: Attach the Faulty Volume to the Rescue Instance #

• Go back to the Volumes page, right-click your detached root volume > Attach Volume.
• Select your rescue instance from the dropdown, and set the device name to something like /dev/xvdf (avoid using /dev/sda1 since that's the rescue instance's own root volume).
• Connect to your rescue instance via SSH.

Step 4: Mount the Volume and Fix Permissions #

Once connected to the rescue instance:

1. Create a mount point for the faulty volume:

   sudo mkdir /mnt/recover
   

2. Mount the volume (replace /dev/xvdf with the device name you chose earlier):

   sudo mount /dev/xvdf /mnt/recover
   

3. Fix the base /home directory permissions (default should be 755—only root can write, others can read/execute):

   sudo chmod 755 /mnt/recover/home
   

4. Fix the user's home directory (e.g., /home/ubuntu—default is 700 so only the user can access it):

   sudo chmod 700 /mnt/recover/home/ubuntu
   sudo chown -R ubuntu:ubuntu /mnt/recover/home/ubuntu
   

5. Fix SSH-related permissions (this is why you couldn't connect—SSH rejects overly permissive .ssh directories):

   sudo chmod 700 /mnt/recover/home/ubuntu/.ssh
   sudo chmod 600 /mnt/recover/home/ubuntu/.ssh/authorized_keys
   

   If you have other users on the instance, repeat steps 4 and 5 for their home directories too.

Step 5: Reattach the Volume to the Original Instance #

• Unmount the volume from the rescue instance:

  sudo umount /mnt/recover
  

• Go back to the Volumes page, detach the volume from the rescue instance, then reattach it to your original instance as the root volume (device name /dev/sda1 or /dev/xvda).

Step 6: Test the Original Instance #

• Start your original EC2 instance and try connecting via SSH—it should work now.
• Double-check other permissions if needed (e.g., any custom files in /home that might need specific permissions).

Quick Notes #

• You can terminate the rescue instance once you're done to avoid unnecessary charges.
• Always double-check device names when attaching/detaching volumes to avoid overwriting the wrong data.
• If you're unsure about default permissions, compare them to the rescue instance's /home directory—they'll match a fresh Ubuntu setup.

内容的提问来源于stack exchange，提问作者Shibdas Kumbhakar