Puppet users - Jan 2009 - Is Anyone Using Puppet to Administer Multiple Clusters?

Hi, Gang...

  The Subject: line says it all pretty much. We''ve
got six clusters spread over a nice sized geographical
area. I wonder whether anyone is using Puppet in a kind
of, let''s say, "cascaded" style to manage the configurations
on all the nodes of all the clusters, given that each cluster
has a dedicated outward facing management node and the usual
mess of compute nodes hidden behind that management node?

  That is, I wonder whether one could have each of the
head nodes act as a puppet master to their respective compute
nodes, while having the head nodes being clients to an overarching
puppet master that can see all of the head nodes?

Cheers,
tomcat


--~--~---------~--~----~------------~-------~--~----~
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+unsubscribe@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/puppet-users?hl=en
-~----------~----~----~----~------~----~------~--~---

I''m sure people are doing this, possibly even in significant numbers 
because it doesn''t seem too complicated. What do you see (if anything) 
as being hurdles?

--~--~---------~--~----~------------~-------~--~----~
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+unsubscribe@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/puppet-users?hl=en
-~----------~----~----~----~------~----~------~--~---

You could easily do the setup you mentioned. Or, if all the dispersed
nodes can access a single server, then you could even simplify with a
single puppet master. Puppet''s updates are client initiated.

-adam

--~--~---------~--~----~------------~-------~--~----~
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+unsubscribe@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/puppet-users?hl=en
-~----------~----~----~----~------~----~------~--~---

On 1/12/09 1:03 PM, Joe McDonagh wrote:
> I''m sure people are doing this, possibly even in significant
numbers
> because it doesn''t seem too complicated. What do you see (if
anything)
> as being hurdles?
Environments of modules are your friend.... :)

Cheers,
Ryan

--~--~---------~--~----~------------~-------~--~----~
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+unsubscribe@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/puppet-users?hl=en
-~----------~----~----~----~------~----~------~--~---

On Mon, Jan 12, 2009 at 1:35 PM, Tony Kocurko <akocurko@mun.ca>
wrote:
> Hi, Gang...
>
>   The Subject: line says it all pretty much. We''ve
> got six clusters spread over a nice sized geographical
> area. I wonder whether anyone is using Puppet in a kind
> of, let''s say, "cascaded" style to manage the
configurations
> on all the nodes of all the clusters, given that each cluster
> has a dedicated outward facing management node and the usual
> mess of compute nodes hidden behind that management node?
>
>   That is, I wonder whether one could have each of the
> head nodes act as a puppet master to their respective compute
> nodes, while having the head nodes being clients to an overarching
> puppet master that can see all of the head nodes?
>
> Cheers,
> tomcat
Not implemented it yet, put I have been looking at doing something
like this. Basiclly my idea was to have a puppetmaster on the head
node dedicated to managing the cluster, and have the main puppetmaster
maintain the puppet manifests on the head node. This would keep all
the configuration files in one place and reduce the outbound/inbound
traffic load from the cluster.

Evan

--~--~---------~--~----~------------~-------~--~----~
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+unsubscribe@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/puppet-users?hl=en
-~----------~----~----~----~------~----~------~--~---