Today I was playing with EC2, trying to launch a RHEL 6.3 instance so that I could then install the latest version of Katello and beat a bit on it… just for fun, you know? Using the EC2 Management Console web interface I used the “classical” wizard to select all the components I wanted for a m1.large instance, making sure to edit the default 7.5 GB root partition they give you so that I could have more space available to synchronize content… but when the instance finally came up I realized that my disk space was still showing the default value:

[root@ip-aa-bb-cc-dd ~]# df -h

Filesystem            Size  Used Avail Use% Mounted on

/dev/xvde1            5.7G  1.7G  4.0G  30% /

none                  3.7G     0  3.7G   0% /dev/shm

I wrecked my brain about that for quite some time, retracing my steps and even terminating my instance and starting from scratch a few times, thinking that perhaps I had missed an obvious step. Eventually I came across a few posts online and was able to solve my problem, which I will try to describe below. Obviously, feel free to read the original posts for more information. The steps are as follows:

1. Once your instance is up and running, stop it selecting the Stop option from the Management Console window.
2. Now, switch to the Elastic Block Store section and select your instance’s volume from the  Volumes subsection.
3. Detach the selected volume.
4. Select the option to create a snapshot off of the detached volume.
5. Switch to the Snapshot subsection and select the newly created snapshot.
6. Select the Create Volume option and create a larger volume.
7. Go back to the Volumes subsction, select the newly create volume and attach it as the root volume for your instance (should be /dev/sda1).
8. Restart your instance.

As soon as the instance is back online, ssh to it and verify that the disk size has not changed:

Now, resize the root partition so that it can “absorb” the larger volume we created:

[root@ip-aa-bb-cc-dd ~]# resize2fs /dev/xvde1

resize2fs 1.41.12 (17-May-2010)

Filesystem at /dev/xvde1 is mounted on /; on-line resizing required

old desc_blocks = 1, new_desc_blocks = 5

Performing an on-line resize of /dev/xvde1 to 20971520 (4k) blocks.

The filesystem on /dev/xvde1 is now 20971520 blocks long.

If everything goes well you should now see a much larger disk available (a 80GB volume in my case):

[root@ip-aa-bb-cc-dd ~]# df -h

Filesystem            Size  Used Avail Use% Mounted on

/dev/xvde1             79G  1.7G   77G   3% /

none                  3.7G     0  3.7G   0% /dev/shm

The following posts were helpful to me:

• https://forums.aws.amazon.com/thread.jspa?messageID=309754&#309754
• http://alestic.com/2010/02/ec2-resize-running-ebs-root

PS: After writing this I wonder if I’m expected to run resize2fs after editing the root partition during the wizard process and starting the instance… will try that next time.