-
Notifications
You must be signed in to change notification settings - Fork 90
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Local account locks itself #183
Comments
Are your users required to change their local passwords on a regular basis? As for information... example super.logs showing password failures is where you should start. |
Hi! |
Hi again! |
According to that log, you are correct that For example here are the types of log lines that indicate None of those lines appear in the log, thus the user account being locked has nothing to do with |
Thank you very much for your clarification and confirmation! The example lines will help to identify if other computers had issues in combination with the SUPER script. In case I should find any, I will report back again. |
Hi! I am sorry to reopen this case, which also might be in context with issue #219. We are still having the issues mentioned. User accounts (auto-locked after 5 failed attempts by password policy) are being locked while the computer is in „sleep“ (lid closed, but apparently on AC and for this reason active) and In In We currently are on 4.0.2, but the respective lines are unchanged in 4.0.3 as well as in 5.0.0beta. So my question is: Is the empty/space password given intentionally for the software download? In asu_workflow_log, we can see the following lines on affected machines: I am unsure if this is the issue or if the account lock was done earlier. But after unlocking the user account ( Sorry for bothering again on this issue, but this causes trouble every time Apple publishes a new update. Kind regards, |
Thanks for your thorough investigation. So the goal is to download the update in the background without authentication before bothering the user to restart. To be clear, It would seem that I should probably defer the download workflow until the computer is awake. |
Thank you for your quick reply and clarification! So actually, you don't see an issue in passing an empty/space stdin string to the As I can see from my own experience, MacBook Pros with certain apps installed (Adobe CC, Google [Chrome] Updater, ...) do not hibernate (for long) even when the lid is closed — as long as AC is still connected. I've seen them pulling 40 watts all over the night while being fully charged. Only when power supply is disconnected, they really fall asleep. We've about 500 Macs managed and about 10% are having this use case as we've now found out. And they appear to be affected by the issue. If it's the lid or (missing) hibernation, that should be doable, but just want to make sure it's not the missing password. — And yes, I've noticed and appreciated the bunch of hacks and special cases you are already handling in |
Significant testing went into the design of Again, downloading macOS updates does NOT require a password. Apple has simply never taken the time to fix the |
Maybe I can illustrate my experiences a bit more based on the attached Within the above But starting with „macOS Sonoma 14.6“ (July 30), the behavior is different: Initial download at 9:01:43, ready at 9:14:50. As the user did not seem to react on the update notification, subsequent download requests are being fired at 9:30:13, 9:46:14, 10:11:15 and so forth. But, after the The very same situation can be seen for „macOS Sonoma 14.6.1“ last week. So I would suspect that Apple might have made I will report back! |
Hi again! Since there are currently no pending Apple updates for me to test, I did another test on another Silicon system in The test system is on macOS 14.5 currently, so 14.6.1 would be suitable. In administrative context, in For the first call, I did not enter my password and simply hit I repeated the same command (without entering my password) and received the same result (skipping the download itself but still including For the third time calling the command, I also gave my account password, resulting in: The only difference so far is that my test system does not have local password policies, so I cannot prove that the counter is being incremented. In case you should also need this test, let me know and I will reinstall/downgrade one of the productive machines. But for now I would suggest to pass the password even for the download – as long as it is known to |
Though I have not had time to confirm this behavior... it would break many |
Are you restricting macOS updates to administrators only? |
No, not that I knew. Otherwise, updates would not work at all since all cached user passwords are for standard (non-admin) accounts. Also on my test system (actually my own), user „ron“ is a standard account. I do definitely understand that altering the way downloads are initiated in |
So... I just ran
|
For good measure... I even ran the line that Also to be clear,
|
Well the good news is that I've replicated the issue.... the bad news is that it's definitely the It appears if that key is deployed then the unauthenticated download does indeed fail. Sigh... I'll add it to the list of exceptions that I need to work around for janky |
Dear Kevin, thank you for your research! I am happy you were able to replicate it though ... I will from my side slightly change the code in the Since you were quoting ¯\_(ツ)_/¯ |
The good news is that even though I didn't wait to show you the final results... it can work... I tried it again just now.
However, the terrible, awful, horrible, bad news is that I've also discovered that it only works if at no point in the history of the system has anything ever attempted to even engage with |
Confirmed 14.5 only. 14.4.1 Does not have this problem. |
Thank you for your investigation! Great to know! |
This is next on my priority list... |
Hi Kevin! Just a brief update: Since macOS 14.7 was released this week, we've had no further lockouts. I indeed was updating and testing the (modified) SUPER release 4.0.2 we were using so far to send the PW along with the user name for downloads, but this was only rolled out on a few test devices. The vast majority of all managed Sonoma Macs (350 of which 10 are test devices) already have the 14.7 update and no user complained about a lockout! So it may be that Apple revised the behavior in 14.6.1 ...?! Kind regards, |
ugh... of course the results are inconsistent... |
That was my approach as well: First try one unauthorized download and check user lock status via
In case of a locked out account, I won't try anyway. Otherwise, if the failed attempt count increases after the first download attempt, I will try one more time giving the password. If the counter still increases, give up. |
My goal is to avoid locking the account entirely by only attempting an unauthenticated download once... then after that assume that authenticated downloads are required for that version of macOS. The complication comes in where there are so many different authentication and workflow options offered by Coding = 1% do a thing, 99% don't do the wrong thing. |
Fixed for local authentication in v5.0.0-beta4... but not for MDM workflows... so keeping it open for now... |
Thank you for your work and clarification(s)! As for A bit offtopic ... but how close is 5.0.0b4 to final? I need to do extensive localization customizations and my customer asks for Sequoia support ...? And another question for which I did not find the answer quickly enough in the Wiki (sorry!): If Major Software Updates are postponed for 90 days in a config profile, will Cheers, Ron. |
The final big push for v5.0.0 completion is almost entirely MDM workflow stuff. As far as changes to I have never tested v4 against macOS Sequioa... I have heard reports that upgrades from macOS 14 work... but I know for a fact that older OSes will not work when upgrading to macOS 15 and that the restart validation on macOS 15 might mistake security updates as system updates. |
Closing this issue with the release of https://github.com/Macjutsu/super/releases/tag/v5.0.0-beta5 So because the macOS download authentication failure is inconsistent, |
Thank you very much! Will keep you updated. |
Hello,
We use a macOS configuration profile that blocks the local account after 5 incorrect password entries.
Since we rolled out Super on the devices, they seem to lock themselves automatically because the password is entered incorrectly in the background by Super?
Users must enter their password to update; no service account should be used on the device.
Anyone else have this problem? What information is needed?
Thanks
The text was updated successfully, but these errors were encountered: