Posted on 10-06-2015 07:25 AM
After looking through the scripts that were red flagged post upgrade to 9.81 it appears that the alert is inaccurate at best and does not have an obvious way to have it recheck.
Posted on 10-06-2015 07:33 AM
Its odd that it would have flagged that script, if in fact the image above is how the script was pre upgrade. But just click the Dismiss button to the far right of the alert section on top. The alert will go away.
Posted on 10-06-2015 07:36 AM
That's weird @mm2270, I never had a "Dismiss" option... I just cloned the script and it went away.
Posted on 10-06-2015 07:41 AM
Huh, really? That is weird. I've seen 2 JSS's now upgraded to either 9.8 or 9.81 and in both cases any alerts about the binary references always had the Dismiss button there we could click to make the message go away. Not sure why it wouldn't be showing up for everyone else's JSS.
Posted on 10-06-2015 07:48 AM
Hey @andrewseago
Thanks for posting this. I'm not sure what the scan saw in the script that thought it needed updating. The scan looks at literal text vs actually executing something and analyzing it's output. We can take a look at it here.
Regarding rescanning - we didn't add a rescan function in. We basically wanted it to scan it once, show where there is potential changes needed, and then stop scanning. The decision was mostly made to not add the expense of scanning on a regular basis. Our thoughts were that changing the path is top of mind for this short time frame, but then won't need to happen again so didn't want to add functionality to continuously scan.
Let us know if there are some other issues we need to look into there. Thanks!
Posted on 10-06-2015 07:52 AM
@john.miller Any ideas on why some folks would not be seeing a "Dismiss" button in the bar with the alert message? I have seen it in every case of a script or policy needing to be updated. I found that I needed to click that even after changing the invalid references. My assumption is because of what you just mentioned, that the JSS does not rescan for the invalid references so there's no way for it to auto clear after the fact. Hence the need for the dismiss button. But if some folks aren't seeing that, how would they get rid of the messages?
Posted on 10-06-2015 08:03 AM
I had one like that where the script wasn't wrong. I dismissed it.
There should have been a "re-check" command/button to make sure you'd gotten everything.
I think these are the MySQL commands they are using. Don't try this if you're not comfortable with the MySQL command-line interface:
select script_id, file_name from scripts where script_contents like "%/usr/sbin/jamf%";
select policy_id, name from policies where run_command like "%/usr/sbin/jamf%";
Posted on 10-06-2015 08:31 AM
Hey @mm2270
Correct. The dismissal is explicit, and would need to get hit even after resolving the issues. We're looking into why some folks aren't seeing it.
Posted on 10-06-2015 08:54 AM
For those of us who have this problem and are somewhat comfortable with hacking, can any give the URL structure of the dismiss button?
Posted on 10-07-2015 09:05 AM
On not seeing the dismiss button:
https://jamfnation.jamfsoftware.com/discussion.html?id=17278
Posted on 10-15-2015 02:19 PM
Has anyone else noticed the notification appear when both /usr/sbin/jamf and /usr/local/bin/jamf appear in the script? That’s a slightly different kind of inaccuracy.
Fwiw, in current testing I’ve seen the “Dismiss” button, but I wasn’t watching out for it specifically.