The Shining Path of Least Resistance

LeastResistance.Net

Spiceweasel 2.0

Posted by mattray on January 14, 2013

One year after the release of Spiceweasel 1.0, Spiceweasel 2.0 is now available! For those of you unfamiliar with Spiceweasel, it is a command-line tool that helps manage loading Chef infrastructure from a manifest file. The JSON or YAML manifest is a simplified representation of the Chef repository contents that you may use to recreate and share how to build the application or infrastructure. This file is validated to ensure that all the components listed are present and the correct versions are available, and can (and should) be managed in version control with the rest of your repository.

There are minor updates to the manifest syntax in 2.0, with a focus on clarity. Spiceweasel has 3 major new features since the 1.0 release :

Execute
Spiceweasel now has the ability to directly execute the knife commands, creating (or deleting or rebuilding) the infrastructure described in the manifest when you use the -e/--execute flag.

Extract
Rather than write your manifest from scratch, Spiceweasel may generate the knife commands or manifests for you. Running spiceweasel --extractlocal generates the knife commands required to upload all the existing cookbooks, roles, environments and data bags found in your Chef repository with validation. spiceweasel --extractjson or spiceweasel --extractyaml will generate the JSON or YAML manifest for your Chef repository, which may be captured and then tracked in version control.

Clusters
Clusters are not a type supported by Chef, this is a logical construct added by Spiceweasel to enable managing sets of infrastructure together. The clusters section is a special case of nodes, where each member of the named cluster in the manifest will be tagged to ensure that the entire cluster may be created in sync (refresh and delete coming soon). The node syntax is the same as that under nodes, the only addition is the cluster name.

clusters:
- mycloud:
  - openstack 1:
      run_list: role[db]
      options: -S mray -i ~/.ssh/local.pem -Ia0b27dd9-e008-4a13-8cda-a2c74a28b7b1 -f 3
  - openstack 2:
      run_list: role[web] recipe[mysql::client]
      options: -S mray -i ~/.ssh/local.pem -Ia0b27dd9-e008-4a13-8cda-a2c74a28b7b1 -f 2
- amazon:
  - ec2 1:
      run_list: role[db]
      options: -S mray -i ~/.ssh/mray.pem -x ubuntu -I ami-8af0f326 -f m1.medium
  - ec2 2:
      run_list: role[web] recipe[mysql::client]
      options: -S mray -i ~/.ssh/mray.pem -x ubuntu -I ami-7000f019 -f m1.small

This would generate the following knife commands:
knife openstack server create -S mray -i ~/.ssh/local.pem -Ia0b27dd9-e008-4a13-8cda-a2c74a28b7b1 -f 3 -j '{"tags":["mycloud+roledb"]}' -r 'role[db]'
knife openstack server create -S mray -i ~/.ssh/local.pem -Ia0b27dd9-e008-4a13-8cda-a2c74a28b7b1 -f 2 -j '{"tags":["mycloud+rolewebrecipemysqlclient"]}' -r 'role[web],recipe[mysql::client]'
knife openstack server create -S mray -i ~/.ssh/local.pem -Ia0b27dd9-e008-4a13-8cda-a2c74a28b7b1 -f 2 -j '{"tags":["mycloud+rolewebrecipemysqlclient"]}' -r 'role[web],recipe[mysql::client]'
knife ec2 server create -S mray -i ~/.ssh/mray.pem -x ubuntu -I ami-8af0f326 -f m1.medium -j '{"tags":["amazon+roledb"]}' -r 'role[db]'
knife ec2 server create -S mray -i ~/.ssh/mray.pem -x ubuntu -I ami-7000f019 -f m1.small -j '{"tags":["amazon+rolewebrecipemysqlclient"]}' -r 'role[web],recipe[mysql::client]'
knife ec2 server create -S mray -i ~/.ssh/mray.pem -x ubuntu -I ami-7000f019 -f m1.small -j '{"tags":["amazon+rolewebrecipemysqlclient"]}' -r 'role[web],recipe[mysql::client]'

What’s important to note here is the use of tags to identify them, we’ll be coming back to this in future releases.

Roadmap
As always the code is available at https://github.com/mattray/spiceweasel with a detailed changelog. The code has been heavily refactored into discrete libraries and the next few releases should add additional cluster features and integration with Berkshelf and Librarian.

Feature suggestions and patches are always welcome, thanks for using Spiceweasel!

About these ads

6 Responses to “Spiceweasel 2.0”

  1. juddmaltin said

    We might want to help Crowbar to ingest Spiceweasel.yaml/json. Could be very interesting.

    • mattray said

      That could work well, since I’m using it as the basis for the Sputnik Cloud Launcher. Plus there’s a pull request to add straight Ruby support that will go in soon.

  2. […] some great posts on the subject in his blog A Fistful of Servers. Matt Ray has supplied “Spiceweasel,” a great tool to get your Chef environment off the ground. But clear articulation of […]

  3. […] if you can’t wait for that for loop to finish? The good news is that there are tools like Spiceweasel that allow you to write a sort of table of contents for your […]

  4. Robert Tsai said

    Hi Matt,

    We are using Spiceweasel pretty heavily with multiple cloud providers and it’s working well. We recently started using ec2 and ran into an issue with spiceweasel –delete. The knife ec2 command needs the –node-name option to delete a node by it’s node name. Is there a way I can pass that in to spiceweasel? (The error below is what I encountered.) Any help would be appreciated.

    ERROR: Fog::Compute::AWS::Error: Malformed => Invalid id: \”host.domain.net\”

    Thanks,
    Robert

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

 
Follow

Get every new post delivered to your Inbox.

%d bloggers like this: