0

i am using kitchen-ec2 with chef-zero provisioner:

$ cat Gemfile | grep kitchen
  gem "test-kitchen", "1.4.2"
  gem "kitchen-ec2", "0.10.0"

$ cat .kitchen.yml | grep require_chef_omnibus
  require_chef_omnibus: true

where the current chef_omnibus version is 12.4.3-1.

before posting a pull request over kitchen-ec2, i would like to verify whether it is possible to synchronize cookbooks from s3 (downloading them from s3 into the ec2 instance), rather then synchronizing the cookbooks by uploading them into each ec2 instance per suite?

synchronizing the cookbooks from the local machine to s3, and then downloading the cookbooks from s3 to the ec2 instance -- is drastically much faster and consumes less communication bandwidth.

Mr.
  • 9,429
  • 13
  • 58
  • 82
  • Have you considered using a chef server? It would provide the workflow you're looking for plus a set of Chef specific APIs for managing your cookbooks and other associated Chef entities. Chef server can be configured to use S3 as it's cookbook storage, but I find it scales plenty without going to that bother. Finally you can start with the hosted service provided by Chef, which already runs on Amazon. – Mark O'Connor Oct 04 '15 at 12:30
  • @MarkO'Connor: indeed, i was asking whether such flow can be avoided :) i also fixed my post, i had `chef-solo`, where it should have been `chef-zero`. sorry about that. – Mr. Oct 04 '15 at 21:14
  • In this case you're using test kitchen. The cookbooks under test are local to the development machine and need to uploaded to the test virtual machine. – Mark O'Connor Oct 04 '15 at 22:40

0 Answers0