High Sierra - OSInstall.mpkg - Install Failures, solved with "Select Startup Disk"

[dpeeri]

161,283 Hits on Ask Different, and it's happened to us on 50% of our computers, and it happens on every point update (10.13.2, 10.13.3, 10.13.4). It's fixable with a "Select Startup Disk" but it puts our users out to dry with an unbootable system if they're not aware of the workaround.

I'm concerned with the fact that this keeps happening on every point update and there's radio silence from both Apple and JAMF. Shouldn't there be a knowledge base or a root cause by now?

https://apple.stackexchange.com/questions/299808/high-sierra-the-path-system-installation-packages-osinstall-mpkg-appears-to-b

Image attached for 10.13.4 , they Error message has changed slightly but it's the same issue.

[Look]

We get this periodically, I reduced the frequency in our environment by having the check for updates occur on startup (restricted to after hours) and force restarting the machines in the middle the of the night so they would check for the update then. This garaunteed there was no logger in user and that the machine was free to restart again immediately if needed.
I am pretty sure it's some function of the installer being put in a cache location or similar and the location being set as the boot location on the device, then the cache or whatever it is being cleared prior to the device eventually restarting. I have had machines do this weeks after we stopped pushing them updates so I know that sometimes the restart fails and then the users will stay logged in for ages afterwards.
In case anyone else needs help with this you can resolve it fairly easily, two ways at the client end.
For an SSD you can option boot and then control select the correct boot volume.
For a fusion drive you need to shift boot into safe mode and use Startup Disk preferences to select the correct boot volume.
I haven't tried it, but I imagine you can probably Recovery Boot and then use Startup Disk from there as well.

[blacktip]

We usually have folks boot into recovery and choose the disk from there. Once rebooted, we have them update in the traditional way. This will ensure it doesn't happen again.

[bigbangerp]

Forgot to mention, we haven't been using JAMF to manage updates for this. We've had users self-update.