[chef] Re: Re: Re: Chef Solo experience / best practices?

[Booker Bense < >]

The primary difference between solo and server is the ability to use search in the cookbooks. For instance, I could search for nodes that have the "dns server" role and use the results of that search to populate /etc/resolv.conf. That's a kind of trivial example, but the ability to search adds an extra layer of

automation. When I was looking at puppet vs chef, the search capabilities and database available in server were the primary deciding points in Chef's favor.

We need that desperately, you may not. 

In the current version, the server is a bit of a "mystery box" to set up and run. It really depends on the scale and scope of where you are starting in your 

deployment. 

This podcast  ( Episode 28 ) might offer some insights. 

http://devopscafe.org/show/?currentPage=2

On Fri, Nov 2, 2012 at 2:06 PM, Jens Skott < " target="_blank"> > wrote:

As a chef newbe I have only read about chef-solo some, what are the advantages of running chef-solo vs the chef server version? 

Sent from my iPad

On 2 nov 2012, at 18:21, Jeremy Voorhis < " target="_blank"> > wrote:

Thanks Jeff,

The talk coincided with a few things I've heard about cookbook distribution (HTTP with ACLs) and chef-solo execution, and there's lots to think about here.

On Fri, Nov 2, 2012 at 7:33 AM, Jeff Blaine < " target="_blank"> > wrote:

On 11/1/2012 8:03 PM, Jeremy Voorhis wrote:

Ohai,

I'd like to know if anyone has had success using chef-solo for a
multi-node production deployment. I've heard and read reports of using
it with MCollective a couple years ago, but haven't heard much since.

Ben Rockwood uses it at Joyent.

http://www.youtube.com/watch?v=he7vxhm6v64