[Pulp-list] pulp supporting automated deployment/test scenarios?

Baird, Josh jbaird at follett.com
Fri May 9 12:15:45 UTC 2014


You don't necessarily need to use the pulp-consumer client on your hosts just to use Pulp repositories.  You can simply drop yum repo definitions down to each host that point to the pulp repos.  We automate and manage these yum repo definitions with Puppet.  When kickstarting, simply point your KS to the pulp-repository containing the distribution packages just like you would if it were a normal non-pulp repository.



-----Original Message-----
From: pulp-list-bounces at redhat.com [mailto:pulp-list-bounces at redhat.com] On Behalf Of David Schmitt
Sent: Friday, May 09, 2014 5:28 AM
To: pulp-list at redhat.com
Subject: [Pulp-list] pulp supporting automated deployment/test scenarios?


I'm currently evaluating pulp as a yum repo backend for a Scientific Linux cluster and several other groups of machines. The setup is puppet and foreman based. Development is done in local vagrant boxes.

Using vagrant, I expect regular turn-over of consumers. I gather that pulp-consumer always has to register with the server first using username/password and it seems to me that unregistering the client is a manual process.

I've got several questions I was unable to answer from reading the

  - Is there a way to use pulp repositories without registration?

  - If not, is there a way to timeout registrations easily, so the
    server is not home to a huge number of zombies?

  - If not, how would kickstarting against a pulp repository work?

  - Is it possible to use duplicate consumer-ids (concurrently)?

  - Will pulp-admin rpm repo copy rpm also remove RPMs that are not
    available at --from-repo-id anymore?

  - Does the RPM provider also synchronize the images/ directory
    required for kickstarting?

Thanks for your time, David

Pulp-list mailing list
Pulp-list at redhat.com

More information about the Pulp-list mailing list