[Puppet Users] Puppet master file serve

On 15/05/12 14:48, David Campos wrote:
Hello all,

I am sure that many of my questions may be answered in other topics or
into the documentation but I can not clear my mind about how to
perform some tasks that I want to accomplish.

To get you into context, I have been developing some puppet
configuration files that deal with deployment of our company systems
and its dependencies. Those environments obviously require an
underlying infrastructure (a tomcat, glassfish or oracle) that is
installed and provided by puppet. In order to accomplish that we did
choose to take advantage of our nexus server to provide our binaries
to puppet and use its hierarchy. We developed some ruby functions
placed at lib folder that download those files and leave them at a
specific location of our machine (we did choose ruby in order to get a
platform independent download strategy).

This worked flawlessly when we were working through puppet apply mode
(our virtual machines were provided by Virtual Box and Vagrant) since
the agent was some kind of server-agent and could perform all actions
by itself.

Recently I have been moving forward a puppet master (there are a lot
of distribution issues in stand-alone agents) that would serve
catalogs and several agents that could be located at either physical
or virtual machines (cloud if possible). With this change I have found
that, as documentation says, functions are executed at master computer
so the files that I need are downloaded at the master... How can I
solve this and at the same time not break stand-alone provisioning?

I have been looking at fileserve.conf in order to fix a specific
folder (I would have to work out how to set that folder at either
Windows and Linux) file serving but I do not know if there is a better
solution... Does puppet provide a best practice for this? An
automagically mounted file provider (like the one of the folder files
at modules).

Thanks.
--
You received this message because you are subscribed to the Google
Groups "Puppet Users" group.
To view this discussion on the web visit
https://groups.google.com/d/msg/puppet-users/-/ziVnFXzbZ7wJ.
To post to this group, send email to [email protected].
To unsubscribe from this group, send email to
[email protected].
For more options, visit this group at
http://groups.google.com/group/puppet-users?hl=en.

On Tuesday, May 15, 2012 4:56:11 PM UTC+2, David Campos wrote:

I knew about that features. When a file is placed a files and referenced
through puppet:///<module>/<file> is automatically served to the agent.
Either through puppet master or apply.

My problem came because we have a Nexus server that serves any external
file (we also try to store there our supported versions of third party
applications and software) and a Jenkins server that performs incremental
builds of our software. When I want to mix that together in order to get an
environment (either for dry-run, production, or testing purposes) I need
to: download my infrastructure from Nexus (or wget the correct files),
install if necessary, get the latest version from my module at Jenkins,
deploy.

The first portion of work could be moved to files but I do not find it a
good practice to version it (remember that I was using a standalone puppet
installation before master) and would be hard to change versions through
environments (Nexus allows that, since is just a proxy). The second part,
download from Jenkins, is not possible since the new build has just come
out from the oven.

That's why I did choose a method to download them from an external http
resource and not from static file serve. Does puppet provide a tool for it?
Should I stick with the fixed files folder, fileserve.conf and so?