So here is the command for registering an EBS backed image based on a particular snapshot:

ec2-register --block-device-mapping /dev/sda1=s<snapshot id>::false
--name "<ami name>" --kernel <kernel id> --ramdisk <ramdisk id>
-K <key file> -C <cert file>

(Note: If you want the volume to persist after termination then you’ll want to change the true in the –block-device-mapping flag to false)

So the key is in the –block-device-mapping flag.  The documentation for Amazon states:

Block Device Mapping Documentation

So it says somewhere in there that you can specify the size of the AMI in that particular flag, so if I want to increase the size of the instance all I would need to do is change the flag to read:

--block-device-mapping /dev/sda1=s<snapshot id>:50:true

or whatever size you want.  Afterwards you’ll need to log on to the server and resize the drive.  You can do this by running the following (assuming you’re using an ext3 root file system, which is what I used in my previous post):

sudo resize2fs /dev/sda1

it’ll resize the drive and you should be good to go.

Share/Save

So here is how I did it:

1) Start Up an Instance

The first thing I do is start up an instance.  Everyone has instructions on how to do this from the command line.  But I’m lazy and find that Amazon’s API tools change regularly which is annoying and makes it difficult to follow the instructions other people have written.  I recommend doing it in the AWS Console.

Go to EC2 in the AWS Console.  Choose AMI and find an image that you like (I tend to lean toward Alestic’s Ubuntu images, but that’s just me).  Once I found an AMI I like I highlight it and then choose “Launch Instance”.  I follow the instructions, making sure that I choose the correct security group (aka firewall) and the correct key pair (you access ec2 servers using a key pair rather than a password).

Once the server is up and running, I update everything, configure it, transfer data onto it, blah blah blah.  The usual things that you do when you are prepping a server.  Once I’m all set I’m ready to make it an EBS backed AMI.

2) Create a New EBS Volume

The next thing I do is create an EBS Volume, again doing it in the AWS Console.  Go to the EC2 section of the AWS Console.  Choose “Volumes” and then “Create Volume”.  You’ll want to make sure of two things:

First that the volume is large enough to hold everything.  I forgot to do that the first time around and *thought* I had to do everything all over again (see next post for how to really do it).  The easiest thing would be to just size it right the first time around.

Second that the volume is in the same availability zone as the instance you started up in step 1.  If you aren’t sure what the availability zone is for your instance, go back to the instance, highlight it, and in the bottom panel you’ll see what zone its in.

3) Attach Volume to Instance

When you are done you’ll want to attach the volume to the instance that you created in step 1.  Once the volume is available, highlight it and choose “Attach Volume”.  You’ll be stepped through attaching it to your server.  If you have only one instance running its pretty straightforward.  If you have multiple ones running, you’ll want to make sure that you’re attaching it to the right place.  You’ll be asked to indicate where you want the attached, I usually do /dev/sdf.  Oh and this process won’t work if the volume and the instance aren’t in the same availability zone, hence my warning in the previous step.

4) Copy the Data from the Server to the Volume

This part of the instructions I found over at Winners Don’t Lose.  This helps you format the volume and copy the data on the current instance over to the volume:

In a root shell on the instance:

# mkfs.ext3 /dev/sdf
# mkdir /mnt/target && mount /dev/sdf /mnt/target
# rsync -avHx / /mnt/target
# rsync -avHx /dev /mnt/target
# sync;sync;sync;sync && umount /mnt/target

The above commands did the following:

formatted the entire volume /dev/sdf as an extended 3 filesystem
created directory /mnt/target and mounted /dev/sdf at /mnt/target
rsync’d the root instance-store filesystem to the ebs volume
synchronized the /dev directory from the instance-store filesystem
flush all pending write ops, and unmount the EBS volume

To be honest these are the only instructions I’ve ever been able to follow that doesn’t produce some error that I have to debug.

5) Detach the Volume

The next step is to detach the volume that you just created from the running instance.  Again you can do this in the AWS console by finding the volume, highlighting it, and choosing detach.  Make sure the you’re detaching the right volume though!

6) Create a Snapshot of the Volume

You’ll want to create a snapshot of the volume that you just created.  You can again do that in the AWS Management console but finding the volume, right clicking on it, and choosing create snapshot.

7) Register your AMI

This is the part that took me forever to figure out.  Mostly because the documentation for the ec2-register command didn’t jive with the tools that i had access too.  The command I used was:

ec2-register --block-device-mapping /dev/sda1=<snapshot id>::true
--name <ami name> --kernel <kernel id> --ramdisk <ram disk id>
-K <your pk file>.pem -C <your cert file>.pem

(note i updated this:  if you want the volume to persist after termination then you’ll want to change the true in the –block-device-mapping flag to false)

You’ll want to make sure that you specify and kernel id and a ram disk id, I usually use the same one that my instance I’m creating this new AMI from was using.  You don’t have to do it, but that would mean that if your AMI users don’t choose one during start up, then the instance won’t work.

Note that you may have your key and cert file set as part of your environment variable.  I have multiple accounts so I don’t, and instead just specify them when I’m running any commands (annoying but I don’t do much from the commandline anyway).

Share/Save

So what do these three things have in common?  Each one is being hosted with Slicehost.

Slicehost is virtual private hosting and a subsidiary of cloud computing provider Rackspace.  Essentially all of their servers are in Rackspace’s cloud.  Slicehost allows you to provision servers, schedule backups, create snapshots, clone servers, etc.  But rather than being charged per hour, you’re charged a flat monthly fee (which can go up if you go over your bandwidth allowance).

So just like any best friend there are things about it that I like and thinks about it I don’t like.  I like how easy it is to clone a server or create a back up.  So if I want to upgrade WordPress for example, rather than dumping the MySQL database, backing up the files and then upgrading, I can take a snapshot of the server and then upgrade knowing that I have an exact replica of the system.  I also like how easy it is to do development work.  If there is a bug fix or something that I want to try out, I can clone the server, work on the changes, and know that it’ll work on my production instance (because essentially I’m using my production instance).

I speak so glowingly of it, what criticisms could I have?  Well what would be great is if I could have the best of both worlds.  I wish that in addition to paying a flat monthly fee, I could have the option to provision for just a few hours.  Maybe I could log into Rackspace’s cloud, see my Slicehost servers and clone them there.  This would help me with handling data loss and working on smaller more urgent bug fixes.  Maybe I’ll get lucky, they’ll read this, and figure out how to make all my dreams come true.

Share/Save

Visual depiction of the rotating Moodle architecture

I have a plan to back up Wheaton’s Learning Management System (LMS) to the cloud on a yearly basis. I talked about the basics of this architecture in my post I <3 Redmine and You Will Too.Essentially each academic year I will rotate a years worth of Moodle into the cloud, only keeping the current year and the previous year.  Here’s a picture to give you a little help in understanding.

So now I need to find a way for people that are not me to access the Moodle that lives in the cloud.  The problem is that every moodle has a config file, and in that config file, Moodle asks you to specify the URL for where your Moodle is located, this is slightly problematic because:

1. Amazon EC2 servers are not given IP addresses, they are given a Public DNS like
2. Technically they can be assigned to an IP address but Elastic IPs cost money when they aren’t being used, and frankly I would rather not have to pay extra if I don’t have to.
3. The people accessing these Moodles can’t be expected to ssh into a server and change the config file.

So I need to make this happen automagically when the server starts up.  I tried using the php variable $_SERVER but I got some weird things.  So my idea is to have a start up script call the public-hostname (using the API amazon provides) and then edit the config file.  That way when the server starts up, Moodle will automagically be available.

Share/Save

I never argued *against* PHP/Python/Perl or even use of Linux/Apache by PaaS vendor. Just noted PaaS users *shouldn’t* care.

I apologized and asked for clarification since it seemed to me that there was an argument about whether or not a PHP/Perl cloud was necessary.

I re-read the article a few times, and on the last go around the only new explanations I could see from it was:

1. if i wrote in php a vendor would deploy in another language
2. he meant that the LAM in LAMP was what didn’t matter  (although that didn’t seem to be explicitly stated)

In the meantime though, I’ll just wait for a response (I sent it late last night and people tend to sleep) rather than possibly misinterpret more.

Share/Save

James’ last question: “Is the ‘open sourceness’ of a programming stack even that important anymore?” is a good one, but orthogonal to the discussion about a LAMP/PHP cloud, in my mind.

So why has this conversation gotten me in a tizzy and forced me to write two blog posts in one week?  Well two reasons:

1. Many web based open source projects are built on the back of P languages (PHP, Perl, and Python)
2. The cloud exists to make technology easier for the masses.

Now before I go any further I will disclose that the first language I learned was PHP.  The second one was Perl.  While I know there are wars that start on listservs about what languages are better, that is not the purpose of this post, so put it out of your head and don’t say anything about how much you think PHP sucks in the comments.

Web based open source projects are built on the back of P languages

Now I could write something about this, but I like lists.  So here is a brief off the cuff list of what open source projects use P languages:

1. MediaWiki (PHP)
2. WordPress (PHP)
3. Bugzilla (Perl)
4. Drupal (PHP)
5. Plone (Python)
6. Moodle (PHP)
7. AWStats (Perl)
8. MovableType (PHP and Perl)
9. Subversion (Python and C)
10. Trac (Python)

Now, I’m not going to argue the merits of how good each of these piece of software are. Again that’s not my point.  But they are popular whether you like it or not.  And to just dismiss them would be foolish.

The cloud exists to make technology easier for the masses

Google Docs makes it easier to get to files you need to edit.  Heroku makes it easier to deploy Ruby on Rails apps.  Amazon EC2 makes it easier for you to deploy a server.  The fall out of the cloud is that technology suddenly becomes much easier to create and deploy.

So what happens when my mom wants to write her own blog.  Well, most likely she heads over to WordPress.com and creates a blog for herself.  And if she needs a little more, then she signs up for the WordPress Premium services.  But what if I want my own blog.  Well right now I have my friend hosting my stuff.  But what if I want it hosted elsewhere and I don’t want to maintain servers (since I do that all day long).  My only option right now is to do what my mom is doing, but the reality is, that’s not what I want.  I want something like Heroku but for PHP.  And if I had it, I could install any of the PHP applications listed above.

And I think this conversation is very pertinent given that Google just announced its open source learning management system written in Python and intended to be deployed in Google App Engine.  This could potentially be a big blow to Moodle who doesn’t have a cloud to turn to.

Conclusion

So yes a PHP based cloud does matter.  And yes it should be built because there are lots of people clamoring to use it (or who have already figured out a way).

And yes this blog post could be more coherent.

Share/Save

I was asked to write a piece on the cloud (I know you’re shocked) in the upcoming issue of Educause Quarterly which will be dedicated to the cloud. If you’re peer reviewing for that particular journal, we will now pause so that you can close the window (the peer review process is double blind) ………………………………………………………………………………………..

Ok good. Now that they’re gone. Here is a link to a first pass at my article:

Defining Cloud Computing

Essentially its a rehash of my presentation at NERCOMP with a bit more depth and significantly fewer audience members asking questions (not that I mind questions…that was my favorite part by far. its just that i didn’t actually get to finish the presentation properly because there were so many questions). The article itself defines the cloud, and then uses videos to illustrate each of the definition’s points.

Questions, comments, and feedback are welcome.

And yes, I was trying to be funny when I used the cloud to deliver an article on the cloud (see web address of article for more details).

Share/Save

My first impression is that survey respondents don’t understand what the cloud is, but that shouldn’t have been a shock to me, most people don’t know what it is.  Admittedly this lack of understanding could be due to the way the survey was structured; but even after a couple of tweaks to the order of sections (I moved Software to the front of the survey and Platforms and Infrastructure to the end), respondents were still having a hard time understanding what was what.  I’ll most likely need to do some clean up on the survey, but I think for now I have a great understanding of what is going on in Higher Education in terms of the cloud.

In the survey I tried to define the three main components of the cloud: Software-as-a-Service, Platform-as-a-Service, and Infrastructure-as-a-Service.  After the survey was completed, I realized that I’ll need to do a better job of defining things for the audience members in my NERCOMP presentation.  Something that I though I could spend 10 minutes on will probably need a solid 20 minutes so that everyone can be on the same page.  I think diagrams and other visual aids might really help people understand what these different components are, and how they correspond to computing they are already using.

My second impression is that institutions are very comfortable with using Software-as-a-Service.  Below is a graph showing SaaS usage among respondents.  Facebook is of course the leader in the SaaS cloud race, with Twitter and Google Docs coming in right behind them.  What I think is the most interesting though is that overall, Google has the highest share of the marketplace.

As far as PaaS and IaaS, most institutions are not using these services.  I’m hesitant to show results from these sections of the survey since so many respondents confused software for platforms or infrastructure.  Once I clean things up, I’ll provide more information.  Suffice to say though, few if any institutions are using Infrastructures or Platforms in the cloud.  Those that are are using Amazon Web Services (for infrastructure) and Google Code (for platform).

More to come later.  And thank you to everyone that took the time to take my survey.  It has been eye opening and hopefully will make me a more informed speaker.

Share/Save

As a companion to this presentation. I’m doing a survey, to see what kind of things institutions are using in the cloud. Its definitely not a comprehensive list of what’s in the cloud or even what’s possible in the cloud, but my presentation is only 40 minutes.

At any rate, if you work in academia and you’re interested in/using cloud computing, please please please take my survey.

http://www.surveymonkey.com/s/cloud-academia

Your help will be greatly appreciated.

Share/Save

The first thing I did was start with an already existing EBS AMI. Right now there aren’t too many available but they are growing.  I chose an Ubuntu 9.04 image since I’m going to install a stack on it, and I’ve already done *most* of it before on Ubuntu.

A quick note: it is possible to create an EBS AMI from an S3 AMI, but I wasn’t able to figure that out (quickly).  Although I did find a script that will do this for you: http://gist.github.com/249915.

Next, I wanted to get the Ubuntu server prepared.  I needed to:

• install mysql — http://www.howtogeek.com/howto/ubuntu/install-mysql-server-5-on-ubuntu/
• install ruby on rails — http://www.hackido.com/2009/04/install-ruby-rails-on-ubuntu-904-jaunty.html (up to and including step 6)
• install zlib — http://www.techsww.com/tutorials/libraries/zlib/installation/installing_zlib_on_ubuntu_linux.php
• install blacklight — http://projectblacklight.org (includes solr and a number of gem dependencies)

I did all of this.  One thing I noticed, Blacklight requires me to use non-standard ports (ie not 22 and 80).  While the server was running, I went into the AWS console and opened ports for its associated security group.  The changes took effect immediately.  I definitely liked that.

Finally, I went in to the AWS Management Console and chose “Create Image (EBS AMI)”.

Then I filled in the information including a description of my image:

and voila! EBS backed AMI.  the one caveat is:

DON’T TERMINATE THE INSTANCE UNTIL THE NEW AMI HAS FINISHED BEING CREATED

because then you’ll loose all the work that you have just done.  doh!

Share/Save