I have a policy to update inventory on all machines set to run once every day, however for some reason, out of 600 machines it has not run on roughly a third of them in over two weeks, I only realised this when my smart groups seemed to be falling out of order.
I'm unsure what could be causing this behaviour, the machines are checking in throughout the day, but they are not getting the trigger to update inventory. I have it set as a re-occuring check-in trigger.
I am going to be mailing support about this, but I wanted to see if anyone else has seen this issue?
Edit: Macbook Airs, JSS 9.72
Hmm, are you sure the policy is not running? Have you checked the policy logs on some of the Macs affected by this to see if its running but just failing, and therefore not submitting inventory? The policy has to complete successfully for it to send up new inventory and I've seen cases of it failing due to some issue with the Macs, or problems with some of the inventory collection settings, for example.
Also, are other policies on the recurring check-in trigger running OK on these systems?
I assume from what you wrote that the policy is set up as using the Recurring Check-in trigger, with a Frequency set to Once per day, and its scoped to All managed devices. That correct?
The policy is definitely not running, the machines are just sitting as pending in the logs.
I'm looking now at other policies with the machines that are failing to run the inventory update, and it seems that.. Yes some are doing other policies, but sparsely.. Definitely something wrong happening in my JSS.
And yes that is how I have the policy set up.
Depending on what trigger you've tied the recon policy to, go to the client an manually fire the Recurring Check-in (Casper9)
sudo jamf policy -trigger "Recurring Check-in"
See what comes back. If the policies run, sounds like the launchdaemon that calls it is missing or broken (go and see if there and loaded - with launchctl).
sudo jamf manage
Will ensure all that stuff is in place.