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.
I specifically don’t want to debate the merits of chef-solo vs chef-server,
but I’d love to hear from those using chef-solo beyond dev/test.
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.
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.
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.
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.
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 actually covers some of them on some of them in the video
above.
Testing and development with Vagrant is a great reason to use and support
chef-solo. Using chef-solo at a larger scale, lets you remove the
dependency on chef-server and have more flexibility with your chef runs,
but you lose the benefits of using knife for provisioning, and chef-server
for inventory and node search.
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.
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.
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.
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.
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.
I would like to know what are the best practices for bootstrap/deployment on ec2/openstack/hpcloud/azure etc. while using chef-solo for config management. Appreciate any pointers. Thanks.
From: Booker Bense bbense@gmail.com
To: chef@lists.opscode.com
Sent: Tuesday, November 6, 2012 7:11 AM
Subject: [chef] Re: Re: Re: Chef Solo experience / best practices?
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.
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.
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.