Closed robatjamf closed 9 months ago
We are not MDM experts here unfortunetlyy but have you seen @dan-snelson's blog post about Jamf Pro enrollments with Tart. Maybe it can help?
I work at jamf and can help! Not an expert... but can help here. I will do some digging and see what I can figure out!
Hey @robatjamf, have you found anything?
sorry been busy with another project, can look at this next week if that is okay.
Totally! I was just going around ope issues and checking status of them.
Okay been testing today and I think I might have figured it out but going to do some more testing. Thanks
Okay figured it out
oci ghcr.io/cirruslabs/macos-sonoma-base:latest
- MDM profile installs but doesn't actually do anything till a restart and notice if there are changes a restart is needed to "refresh" the OS. Something is up there...
oci ghcr.io/cirruslabs/macos-sonoma-vanilla:latest
- MDM installed and works normally, will just use this image
So the issue that I have is with the base image. I can compare the packer templates and see if there is something interfering with MDM within the base image
One major difference between the images is that macos-sonoma-base
doesn't have recovery partition. Do you think it can interfere?
Interesting. I don't think that this would be an issue but I could test it... do you set that within your packet template? If so can fork and try my own version.
Thanks
I'm sorry I was wrong. Both images are having the recovery partition removed to save space when pushing images remotely.
checking it...
Closing because it seems not a problem of Tart as VMM but rather that a VM is not compatible with MDM because it misses recovery partition.
MDM Install failure till restart
When using the (osversion)-base and enrolling that device into an MDM server like Jamf Pro, there is a stall, where no profiles come down... some sort of detached sync. To get the profiles to download a restart needs to be restarted for an update.
Hard to test enrollment workflows or MDM workflows with this bug...