Information related chef

Hi,

I am sorry if I am asking a silly question but need some more information on this as we are planning the similar change.

Chef can automate Infra deployment in which we can checkin cookbooks in Git and deploy those cookbooks using Jenkins.
Could you please tell me how we can use this for CI where developers are developing their code for certain software and deploy it on DEV,UAT and Prod server.

I am trying to understand where exactly chef stands. Any Architectural diagram along with workflow will be really helpful to understand so that we can try to setup the same for development teams.
Thanks in advance.

Thanks and Regards

So what I’ve typically seen done is: prod, uat, and dev use separate Chef environments. Jenkins/Bamboo/whatever CI tool you use takes git checkins and uploads the new cookbook versions to the server using Berkshelf. Each non-dev environment then has a set of cookbook pins so they are not impacted by new cookbooks being uploaded. Dev runs without pins and just relies on the dependency solver in Chef to get the most up to date cookbook versions that will play together. Once changes are tested in dev, then the pins are updated in UAT, and then promoted to production during your scheduled maintenance/patching/update/release cycle. Since roles and databags are not versioned, if those are a potential issue, swap out environments for organizations and operate that way instead.

If there are enough people working on Chef code and the risk of testing related changes is too high to impact the development environment, what I’ve done before is create a second organization. That organization is hooked up to your CI system through a separate branch (say “develop” versus “master”), and pushes updates to that organization (unless you skip that whole step and just have the devs upload directly to the test organization), where you have a set of test nodes to work with. On the other hand, if you’re doing strong enough testing locally with Vagrant / Test-Kitchen, a dedicated test organization may not be needed. (At the time it was faster to spin up a new VMWare VM than it was to launch a local VM and Docker didn’t work on Macs well with the boot2docker kludge… things are definitely different now and local development and testing is a lot easier to implement.)

Nathan Clemons
DevOps Engineer
Moxie Cloud Services (MCS)

O +1.425.467.5075
M +1.360.861.6291
E nclemons@gomoxie.com
W www.gomoxie.comhttp://www.gomoxie.com/

Thanks for the information Nathan.