Posted on 08-07-2015 07:46 AM
We recently had to restart the machine running our JSS (on 9.73), and I've noticed that when checking my inventory, the check-in times have not updated since restarting. Everything seems to still be working fine, but I haven't had any issues reported yet. Anyone know what might have gone wrong?
Posted on 08-07-2015 08:13 AM
Check and see if the Tomcat service is running.
And from a client, run jamf checkJSSConnection to see if it can connect or not
Posted on 08-07-2015 08:34 AM
I would think Tomcat would have to be running to view anything in the browser. Its essential to the JSS functioning at all, so I doubt Tomcat is down. You'll probably know if that were the case.
Keep in mind that typically, Inventory doesn't happen that frequently. What is the execution frequency for the normal recon (Inventory collection) set to in your JSS? If its set for once a day for example, its not uncommon for the last inventory time to be a little old.
I'd be more concerned if your check-in times were getting stale, since check-in usually happens much more frequently. Even in a large environment like ours where we've set check-in to occur every 30 minutes + 15 minute randomization, we still see check-ins happening all the time due to the randomization shift.
Posted on 08-07-2015 11:03 AM
We have our machines set to check in every 60 minutes. Normally I wouldn't be too concerned, but all machines stopped checking in 2 days ago.
Also, trying jamf checkJSSconnection shows everything is good.
Posted on 08-07-2015 11:20 AM
First, if you have access to one of the Macs that isn't checking in, look in the jamf.log, since the check-ins should show there.
Second, since the check-in is controlled by a LaunchDaemon, the next thing I would do is check to make sure the LaunchDaemon didn't somehow get unloaded on your Macs. As unlikely as it seems for all Macs to stop running the Daemon, I would at least rule that out. Keep in mind that the Mac checks in to the JSS, not the other way around. IOW, its client --> server, not server --> client.
On a Mac that is no longer checking in or submitting inventory, run:
sudo launchctl list | grep jamf
You should get output similar to this:
- 0 com.jamfsoftware.task.Every 60 Minutes 116 - com.jamfsoftware.jamf.daemon
The task.Every 60 Minutes one is the one you should be looking for. If its not there at all, or the 2nd column isn't showing up with a 0 status, it means its not loaded, or there is some error.
If its loaded correctly, then, well, not sure, may be something up with your server. You didn't recently upgrade it or make any changes on it did you?
Posted on 08-07-2015 11:29 AM
@mm2270 Just checked jamf.log, everything seems normal.
Checked to make sure that it is still submitting inventory, and it does have similar output.
As for changes on the server, other than the restart (which lines up on the day our JSS seemed to have stopped updating check ins), there hasn't been any major changes. I updated our GSX account password in the JSS, but that should be it. Granted, I don't have full access to our server, but I would have known if there was an update.
Posted on 08-07-2015 11:45 AM
OK, and what about the LaunchDaemon? Present and accounted for?
If so, have you done a
sudo jamf recon on a Mac that hasn't checked in, or even just a
sudo jamf log (should update the last check in time) and then check the Mac's JSS record to see if any of those fields are being updated? I'm assuming maybe you tried these steps already, but just checking. I'll admit, your situation is kind of strange. If the processes are running correctly and your JSS is up, there's no reason I can really think of why they would have stopped updating on the server.
Posted on 08-07-2015 01:31 PM
Very weird indeed. I even just enrolled a new mac after imaging, shows the enrollment time as being accurate. However, it shows the last check-in date of 8/4.
Anyways, just tried sudo jamf recon and jamf log, went through without issue; however, in the JSS it still shows last check in date as 8/5. Everything seems to be functional, but its almost like the communication back to the JSS isn't there.
EDIT: And yes, the LaunchDaemon was there.
Posted on 08-09-2015 04:25 AM
@ajh11g very odd.
You mentioned you don't have full access to the JSS, have you spoken to those that do?
Posted on 08-10-2015 09:27 AM
All, we figured out the issue. Our JSS is hosted in a VM which apparently ran out of allocated space. After upping the storage space, everything appears to be working.
Posted on 08-10-2015 09:53 AM
Ah, wow. Glad you figured that out. I'm surprised some other more glaring symptoms hadn't shown up along with this, like, not being able to log into your JSS in the first place.
Posted on 08-10-2015 10:35 AM
@mm2270 Yeah, of course that happened today when I got in; couldn't even get to the login window. It was after that point the other team looked and noticed that the storage was full up.