Deleted machines remain in schedule's target history
We tend to reuse a lot of computer names. I have found that when you remove a computer from PDQ that does not delete the schedule history, therefore some packages that I intended to be installed on a fresh deployment do not get triggered. I may be missing something but is there a way to:
- Remove all computer names from all schedule history that are not currently objects in PDQ?
- Remove the computer's deploy history from schedules when that object is deleted from pdq
0
Comments
The answer is no. However, I feel like the deployment "issue" can be fix if collections and target list are setup correctly. From the post, it seems like you are putting individual host name in schedules. I personally only do this if I plan to deploy one time to one device. My suggested method is to attach a collection or target list to it. That way PDQ will see what is missing from the devices, rather than, this device need these software.
Example 1 - Hostname
Schedule - Heartbeat install Adobe Reader if missing.
Target - MyPC1
MyPC1 will come online and install Adobe Reader. The schedule installed Adobe Reader on MyPC1 and is now finished. Take device out of domain, add a new device with same name and it will be missing Adobe Reader now without reinstalling.
Example 2 - Collections
PDQ Inventory - Create two collections. Devices with Adobe Reader installed and devices without Adobe Reader installed.
Schedule - Heartbeat install Adobe Reader if missing.
Target - PDQ Inventory Collection - Adobe Reader Missing.
MyPC1 does not have Adobe Reader installed and will populate in the missing collection for PDQ Inventory. Once detected by heartbeat, it will install Adobe Reader and update the device to the installed collection. Take device out of domain, add a new device with the same name. PDQ Inventory will see the same hostname, do a new scan, see that Adobe Reader is now missing and add that to the missing collection. Deploy will see this new device in the Adobe Reader missing collection and deploy to the same hostname again.
Let me know if you need more details or if something does not make sense. Hope this helps.
Thanks for the reply. I understand what your going for there, i'll just have to create a ton of collections. But that does appear to be the only way. I am linking my schedules to collections now, but just not that refined; but are collections of desktops or laptops(not going deep enough to see software is or isn't installed). Some of my schedules have 15+ custom packages linked to them. Thanks again!
I did add a ton of collections when building this out. It is well worth it in the long run. Helps future proof and automation is a lot easier. For the 15+ packages, I would add them into a nested package, then add the nested package to the schedule. That way if you need to update one step or package from the group of 15, you can deploy the single updated package to devices that only need that one package and deploy the nested package to new devices to get all 15. Cuts down on wasted bandwidth and makes deployment a bit quicker not having to deploy the other 14 in the schedule. I have individual schedules for each Chrome, Firefox, Reader and Flash. They will automatically update on heartbeat to a "Pilot" or testing group of machines. There is another schedule with Chrome, Firefox, Reader and Flash combined. I deploy this to all the other devices at the end of the month. Collections, conditions and schedules add a lot of flexibility.