Puppetlabs/Splunk Module Help!

46 views
Skip to first unread message

Joseph Karns

unread,
Feb 15, 2016, 2:55:39 PM2/15/16
to Puppet Users
Ok...so I'm trying to get the puppetlabs/splunk module to work with Roles and Profiles via Hiera. This is what I have.

hieradata file
splunk::params::version: '6.3.3'
splunk::params::build: 'f44afce176d0'
splunk::params::src_root: 'puppet:///software/splunk'   <- Mapped on Puppetmaster using fileserver.yaml and yes all the files (rpm's in my case) are there.


oit_splunk.pp
class roles::oit_splunk {
  include roles::linuxbase
  include profiles::ulimit
  include profiles::splunk
}


splunk.pp
class profiles::splunk {
  include splunk::params
  include splunk
}


When I run the agent it picks up nothing....I'm super confused and the module needs WAY MORE documentation. I've added --debug and can see that the class is being applied because I can see the providers are being picked up from the class. Any help would be greatly appreciated.

Thanks
Joey

jcbollinger

unread,
Feb 16, 2016, 9:14:03 AM2/16/16
to Puppet Users


On Monday, February 15, 2016 at 1:55:39 PM UTC-6, Joseph Karns wrote:

When I run the agent it picks up nothing....I'm super confused and the module needs WAY MORE documentation. I've added --debug and can see that the class is being applied because I can see the providers are being picked up from the class. Any help would be greatly appreciated.



I don't immediately see anything wrong with the classes and data you presented.  I observe, however, that seeing log messages about resource providers being identified tells you only that instances of the corresponding resource type are present in the catalog.  That such messages report success in no way implies that any resources of those type, or the classes containing them, are successfully applied.

Even if the run succeeds overall with no resource failures, that in itself does not tell you anything about the details of any resource that was applied.  Look for the debug messages that pertain to the application of the splunk classes and resources.  You might also turn on the --detailed-exitcodes and --no-daemonize options, and test the exit status of the run to see whether it indicates any resource failures.  If not by those means, then how are you in fact determining that the agent "picks up nothing"?  What do you expect to see that you do not see?

I agree with you that the module documentation is lacking.  The manifests themselves are reasonably well commented, however, and of course you can always resort to reading the actual code.


John

Reply all
Reply to author
Forward
0 new messages