Connection timed out for AWS ubuntu cloud instance after ufw disable and enable

Question

I configured AWS EC2 ubuntu 12.04 instance and configured it as a web server. I successfully tested the webserver from my local maching using the Elastic IP.

While trying to install cpanel, there was a step which asked to disabled the firewall(CentOS in the tutorial), I searched the web and used the following command to disable it on Ubuntu

$ sudo ufw disable

Before a server restart, I re-enabled the firewall with eh command

$ sudo ufw enable

This command did showed me that it will change the behavior for the SSH connections. I agreed and pressed Y.

I restarted the server and when I tried to connect to the instance with the following command

$ sudo ssh -i key.pem [email protected]

It gave me this error

ssh: connect to host xx.xxx.xxx.xx port 22: Connection timed out

I tried to search the amazon FAQ page, but there was no ticket on this. Can anyone help me resolve this problem?

Answer 1

Today I ran into the exactly same situation. Here's how I fixed it in 3 easy steps :

Warning : Make sure your volumes are EBS, otherwise you will loose data in it.

Suppose INSTANCE-BROKEN is screwed up.

1.) PREPARE :

Create a new EC2 instance INSTANCE-FIXER.

IMPORTANT : This EC2 instance should be created in the same region (It was "us-west-1b" in my case) that has INSTANCE-BROKEN

Stop INSTANCE-BROKEN. Detach the volume from INSTANCE-BROKEN and attach the volume to INSTANCE-FIXER

2.) FIX :

Now start INSTANCE-FIXER and run fdisk to see what available volumes you have there :

    sudo fdisk -l

    Disk /dev/xvda1: 8589 MB, 8589934592 bytes
    255 heads, 63 sectors/track, 1044 cylinders, total 16777216 sectors
    Units = sectors of 1 * 512 = 512 bytes
    Sector size (logical/physical): 512 bytes / 512 bytes
    I/O size (minimum/optimal): 512 bytes / 512 bytes
    Disk identifier: 0x00000000

    Disk /dev/xvda1 doesn't contain a valid partition table

    Disk /dev/xvdf: 8589 MB, 8589934592 bytes
    255 heads, 63 sectors/track, 1044 cylinders, total 16777216 sectors
    Units = sectors of 1 * 512 = 512 bytes
    Sector size (logical/physical): 512 bytes / 512 bytes
    I/O size (minimum/optimal): 512 bytes / 512 bytes
    Disk identifier: 0x00000000

    Disk /dev/xvdf doesn't contain a valid partition table

Mount the volume and edit this file as shown :

    sudo mount /dev/xvdf /mnt
    sudo vi /mnt/etc/ufw/ufw.conf

Edit and make sure ENABLED=no in the file.

unmount the device "/dev/xvdf" from INSTANCE-FIXER

detach the volume from INSTANCE-FIXER

3.) DONE :

Attach at "/dev/sda1" --- *IMPORTANT - ROOT device, otherwise your instance won't start

to the INSTANCE-BROKEN

Start the INSTANCE-BROKEN, now it should work; and

Terminate the INSTANCE-FIXER.

Now you can Login with SSH.

Answer 2

This happened to me as well. Luckily, I was working on 2 terminal windows, and I didn't get booted off one of the sessions. What I did is I allowed incoming traffic on the default ssh port(22). This worked for me:

sudo iptables -A INPUT -p tcp --dport ssh -j ACCEPT