Scalr now integrates with Chef, the configuration management tool offered by Opscode.

Chef allows you to manage software on given servers. Need to update Apache to a newer version? You bet. Need to apply some security patches? No problemo. It requires the ability to write Ruby code (but who doesn’t know that these days?).

Scalr and Chef together—that’s like peanut butter and jelly.

But Chef by itself doesn’t do the provisioning—you know, actually making the API calls to your cloud infrastructure provider to instantiate servers, create block storage volumes, or even assign IP addresses. It can’t provide the exception handling either (e.g. on server failure, do X then Y). It won’t automatically spin up a new slave when an existing one dies.

You see where we’re going with this, don’t you?

Combine the forces of Chef and Scalr, and you can now use Scalr to create a farm, add a few “base” roles, assign runlists to them, then get Scalr to make the necessary calls to launch hundreds of servers at a time. When the instances boot, Scalr makes sure the assigned runlists are executed. In turn, Chef follows the recipes in your runlist, adding software ingredients and putting the servers in the state you want.

Scalr then becomes a mission control center where you can see your infrastructure, track spending, perform updates, and more.

Set up your Chef server to Scalr from Roles in the menu

Sounds good, right? To get started, you’ll need to (1) have a Chef server running or be using hosted Chef and (2) be on the IPO plan or higher. When logged in to Scalr, go to “Chef servers” under Roles, and add the credentials to your Chef server.

Provided they have Chef on them, your roles will have a Chef side-tab. From there you can configure your runlists, which will run every time a new instance of that role is provisioned.

Assign Chef runlists for Scalr to execute on instance start

We already have a few clients using this, and they are delighted. If you are a Chef user, try it out and let us know what you think. And if you have suggestions for improvement, we’re all ears—as usual.

Cheers,
The Scalr “Bork Bork Bork” Team

Starting March 1, Scalr will discontinue the development plan. Any users still on the development plan must upgrade to a new plan by March 1 to continue using Scalr.

In 2010, we created this free model so that users still developing their applications could access the Scalr service. Scalr has grown a lot since then, and we know our users have too. By upgrading now, our users will benefit from Scalr’s expanded service. In the past month alone, Scalr has rolled out a multitude of new features.

To upgrade to a new plan, click here.

Cheers,
The Scalr Money Countin’ Team

What’s that? You want *another* new feature? OK!

Scalr now allows you to clone your farms with the same roles and settings. This is one of my favorite features because it’s a great example of the operational agility that Scalr affords you.

You know how object-oriented programming allowed us to avoid repeating ourselves in our code (aka DRY, or Don’t Repeat Yourself)? In the same vein, this object-oriented model can also be applied to sets of machines, or clusters—known as server farms in Scalr. Define a model, or blueprint, for some infrastructure, and you can instantiate it into real infrastructure to run your application. And then you can use that same model to deploy your application in China, Italy, or the Great Nation of Kazakhstan, which allows you to avoid repeating yourself.

The model can even be applied to software configuration (i.e. what’s on the machine? WordPress? Postgres? Redis?) as well as the passwords, endpoints, and security configuration of each. Software like Chef and Puppet fill this role and allow you to do this through code.

Following this model, you can now clone your farms to create a copy. Here’s how:

On the Farm View page, select Options > Clone. This will create a copy of your Scalr configuration (note: no data is cloned yet), including all scaling settings, scripting settings, etc. In addition, new EBS volumes, ELB, and Elastic IPs will turn up as if you had just created a new farm.

From your farm's options, select 'Clone'

What’s next? You tell us! One thing that might be useful is cloning the farm’s data too, by creating snapshots of the volumes used in the origin farm. Although you can’t do this from Amazon to Rackspace (because the images are incompatible), it might still be useful to do this between regions of a cloud.

This is a first step towards FarmTemplates: farms that you can create, clone, and share with others, for WordPress, Drupal, and other popular software.

Let’s look at three immediate use cases:

1. You can test new features or a new upgrade before going live.
Want to test the effect of MySQL 5.6 instead of 5.1? Create a clone in your dev environment, modify the clone to update MySQL to 5.6, and test it. If it works, you’re safe to repeat on your production environment. If not, you can solve the problem without taking the site down—or better yet, without your manager yelling at you.
2. Create backup with some fancy copy-paste.
Deploy your farm clone to another datacenter, so that if the first goes down you can still operate.
3. Create copies for your clients
If you repeatedly create farms for your clients, internal development groups, or more, then you can create a master farm and clone it instead of repeating yourself.

We think the new cloning feature will find a special place in your heart, especially because it’ll help you save time (and your ass). You’ll never have to repeat yourself again! Yourself again!

We are delighted to announce public support for MongoDB in true Scalr fashion: self-managed, automatically scaled, and easily configurable.

Mongo takes its name from the adjective “humongous,” and it’s a NoSQL database designed for scalability and developer agility. Big names like foursquare, Intuit, and Shutterfly (among others) all use Mongo.

Up until now, our users had to create their own Mongo roles, scale them manually, and write their own scripts to manage clusters of Mongo instances. You had to dig into the Mongo docs to figure out whether to add a replica set or a new shard, dig deeper to find out how to do so, and even deeper to find out the best way of doing it. And that doesn’t even cover the complexity of scaling down.

Check mongo status from your farm's options

Scalr feels your pain, and that’s why we’re rolling out this new automated role for Mongo. Need more write capacity? Click to add a new shard. Need more read capacity? Click to add a new replica set.

Add shards and replica sets with ease!

While you’re playing around with Mongo, we’d love to hear feedback from our users. Email us at suggestions@scalr.com to let us know how the process works for you. In the meantime, we’re working on adding configuration presets for it.

Ready to get this party sharded? Start by creating a new farm at https://my.scalr.net/#/farms/build and adding Mongo found under the database category.

Whip out yer plastic—Scalr can now accept credit card payments.

After an arduous, tedious, and laborious process, Scalr beat the first PCI DSS boss and leveled up to a level 1 PCI compliant company. Next up: saving the princess.