Issue #4240 has been updated by Bruce Richardson.
Brian Gallew wrote: > Nagios being picky is actually a very good motivator for this feature: you > don't want a typo to crash Nagios. > Well, maybe (the generic) you do, but *I* don't. I would much rather have a > typo create a new group which will then show up in Nagios. I really do not see that as good practice; Changes should be tested before being pushed out; errors should not be hidden and allowed to have arbitrary side effects which may not be detected until somebody actively goes looking. If you want to be sure that you never name, in a nagios_host's hostgroups, a group which does not exist, use dependencies; that way, your errors will be caught when the configuration for the node is compiled, with no threat to Nagios. It's quite easy to write a simple set of defines which will automatically create dependencies between the different naginator resource types. If you want Naginator to help you avoid error, then it would make more sense to have Naginator set up those dependencies for you automatically. I could get behind a ticket like that quite happily. In Nagios, hostgroups can have more properties than just their membership; removing the resource would make it impossible to set those. I think this would be a retrograde step and make Naginator less useful to a lot of people. Besides, having the choice of making group memberships specified in the host_groups or in in a hosts's hostgroups field can have an important impact on scalability; forcing this choice onto people can force performance problems on people. > > The current requirement for explicit hostgroup creation means that certain > natural methods of group creation (e.g. hardware/OS combinations, perhaps > with versions) cause Nagios to crash every time you start using a new OS, > platform, or perhaps even with an upgrade. As I said, use dependencies. I make sure that all my services depends on the hosts or hostgroups they check, that hosts depend on the groups in their hostgroups field, that host_groups depend on the hosts in their members field and so on. > One thing about 0.25.x is that you have this beautiful inline_template() > function, but you can't use it for anything other than an rvalue. Hopefully > the "Ruby in manifests" feature will make dynamic manifests a reality. inline_template isn't beautiful; it's a hack. ---------------------------------------- Feature #4240: nagios_hostgroups should be removed https://projects.puppetlabs.com/issues/4240 Author: Brian Gallew Status: Needs more information Priority: Normal Assignee: Category: Target version: Keywords: Branch: The nagios_hostgroups type is redundant. Hostgroups are declared as elements of nagios_hosts. It should be easy enough to generate the hostgroups by scanning through the hosts and collecting all the unique entries. -- You have received this notification because you have either subscribed to it, or are involved in it. To change your notification preferences, please click here: http://projects.puppetlabs.com/my/account -- You received this message because you are subscribed to the Google Groups "Puppet Bugs" group. 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-bugs?hl=en.
