Hi Jed,
On Feb 8, 4:54 pm, Jed <jedbl...@gmail.com> wrote:
> Hi All,
>
> I want to have a bank of puppet servers that will be be fronted by a
> load balancer(anything such as apache proxy, or passenger..etc), these
> server banks will be the puppet masters for our internal dev/stg/prod
> environments/servers, as well as our EC2 cloud instances
>
> so i will have 4 total separate environments...
I have a Puppet Master cluster load balanced using ldirectord (search
for posts by me from about a month ago). There was also a very
detailed post on how to achieve the balancing with Apache / Passenger
recently (2-3 weeks ago).
> the core question i have here is that our nagios server in dev has
> different config files then our nagios server in prod ( ie. apache
> conf.d, etc..)
> How can i have puppet recognize the different environments and deploy
> designated config bases on the identified environment.
Environments have separate 'manifest' and 'modulepath', so use
separate site.pp and modules for each environment you have. This will
result in a lot of replicated data, but it will give you perfect
segregation between Dev and Prod.
For example in my own environment:
/etc/puppet/puppet.conf:
...
[production]
manifest = /etc/puppet/conf/production/manifests/site.pp
modulepath = /etc/puppet/conf/production/modules
[development]
manifest = /etc/puppet/conf/development/manifests/site.pp
modulepath = /etc/puppet/conf/development/modules
For 99% of my modules, the Dev tree is exactly the same as the Prod
tree, however if I want to test a change like a new module, or modify
a configuration file, I can just modify the Dev tree's files and copy
them to Production when they pass testing. For your example, managing
Nagios, lets say you have a Puppet module called 'nagios' you'd have a
file system like:
development/modules/nagios/manifests/init.pp
development/modules/nagios/files/somefile
production/modules/nagios/manifests/init.pp
production/modules/nagios/files/somefile
When you've got something like this in a class:
file { "/etc/nagios/somefile":
source => "puppet:///modules/nagios/somefile",
}
The "somefile" file will come from a the Dev module when your Puppet
clients runs with "environment = development", and from your Prod
module when "environment = production".
Some people use multiple check outs of the one SVN check out, some use
physically separate trees and rsync between to migrate changes from
Dev -> Staging -> Prod. The second way sounds like it'd suit you
better as your Dev and Prod are different a lot of the time.
Hope that helps,
-Luke
> possibly by domain, since our structure is as follows -- even if I
> went this route - i still am at the roadblock of how do I setup puppet
> to do this....
>
> From what i gather, all the nagios servers across all the domain would
> get the nagios rpms i've built using puppet, and then i would have all
> the custom config file sitting on the puppet masters repo(puppet
> master local file system).
>
> poller1.qprvt.tech.com would get abc configs which are designated as
> dev config files for nagios
> poller1.sprvt.tech.com would get xyz configs which are designated as
> staging config files for nagios
> etc...
> etc...
>
> dev = qprvt.tech.com
> stg = sprvt,tech.com
> prod = prvt.tech.com
> ec2 = ec2.tech.com
>
> sorry for the long winded messages.... just trying to convey my
> thoughts...
>
> Thanks for all the help guy/gals -- i'm very excited about puppet --
> but its proving elusive to nail down(at least for me at the moment).
--
You received this message because you are subscribed to the Google Groups
"Puppet Users" group.
To post to this group, send email to puppet-users@googlegroups.com.
To unsubscribe from this group, send email to
puppet-users+unsubscr...@googlegroups.com.
For more options, visit this group at
http://groups.google.com/group/puppet-users?hl=en.
