On 01/21/2014 07:41 AM, Dirk Heinrichs wrote:
> Opposed to what Felix is thinking, my previous message was meant
I thought so, but have issues with the notion that /etc/hosts is to be
ignored in favor of DNS in any production environment.
> seriously. I don't see a trivial solution to your problem that involves
> Puppet and /etc/hosts. You'd need to:
Agreed.
> 1) Have the new node update the hosts file that Puppet delivers. What is
> two noeds come up at the same time? You'd need proper locking.
I don't know how that module works. However, in vanilla puppet, each
machine would export a host{} resource to be collected by all peers.
> 2) Puppet agents only run every 30 minutes, so you'll have a delay until
> all other nodes know the new one.
More or less. There are ways to work around this limitation, e.g. agent
orchestraction using mco etc.
> 3) What happens when a node goes down?
Good point, I'm not sure about the purging of stale exports.
In the case of planned shut down, the node could have a catalog compiled
without the export, so that puppet would take care of cleanup.
> OTOH, what you want to do can be achieved with DHCP and DNS, such that
> the DHCP server updates DNS when a new client comes up.
That, or a custom (peer-to-peer) host file management harness (which is
likely the worse choice).
I concur that this is not in puppet's ball park.
Cheers,
Felix
--
You received this message because you are subscribed to the Google Groups
"Puppet Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To view this discussion on the web visit
https://groups.google.com/d/msgid/puppet-users/52DE6357.4070707%40alumni.tu-berlin.de.
For more options, visit https://groups.google.com/groups/opt_out.
