simple file resource taking longer than it should

[dkoleary]

Hi;

A follow-on to another thread; I'm trying to track down performance issues in open source puppet server (and clients) ver 4.5.0

Reading through last_run_summary.yaml, the vast majority of time is spent on file resources.  I then hit the last_run_report.yaml and, 

as an example, one of the simplest file resources took 17 seconds (assuming I'm reading this correctly):

  File[/etc/ntp/step-tickers]: !ruby/object:Puppet::Resource::Status
    title: "/etc/ntp/step-tickers"
    file: "/etc/puppetlabs/code/environments/dkoleary/modules/mpintp/manifests/init.pp"
    line: 67
    resource: File[/etc/ntp/step-tickers]
    resource_type: File
    containment_path:
    - Stage[main]
    - Mpintp
    - File[/etc/ntp/step-tickers]
    evaluation_time: 17.430422487
    tags:
    - file
    - class
    - mpintp
    - profiles::mpibase
    - profiles
    - mpibase
    - roles::mpibase
    - roles
    time: '2017-01-13T13:08:38.619245800-06:00'
    failed: false
    changed: false
    out_of_sync: false
    skipped: false
    change_count: 0
    out_of_sync_count: 0
    events: []

It'd be hard to get more simple than this file resource:

  file { '/etc/ntp/step-tickers':
    ensure => 'present',
    owner  => 'root',
    group  => 'root',
    mode   => '0644',
    source => 'puppet:///modules/mpintp/step-tickers',
    notify => Service['ntpd'],
  }

The rest of the module is equally as simple.  one package, one service, and one other file resource.  

And, of course, the performance issue is sporadic.  The very next agent run finished the whole thing in less than a second.  I don't believe it's a system resource contention.  *This* box is a linux desktop that was all but idle for both runs.  I'm seeing similar issues on other systems, though, so I'd like to be able to track this down.

Anyone have any ideas or tips on where I can look next?

Thanks for your time.

Doug O'Leary