Activity Monitor: "jamfAgent (Not Responding)"...Maverics and JSS 8.73

[donmontalvo]

We are testing Mavericks 10/9 (13A603), JSS 8.73. Occasionally we see the jamfAgent process freeze, is anyone else seeing this?

external image link

I did a search through the forum, came up dry. I set tag for this thread to jamf binary but this is for jamfAgent process.

After a reboot the jamfAgent process was back to normal. Checked again about an hour later, shows not responding again.

TIA
Don

[charlesludwick · Accepted solution]

We are seeing this and reported it. This is what we got back from Kubley.

Thank you for reaching out about this. This issue is actually a defect in Mavericks currently, as we have even seen our Parallels VM’s report this same way. This issue was opened as DTS 600706615 which ended up in Apple confirming that it is their issue. It is now living as bug #15424012 at Apple. The status of the agent should be harmless, but it might be a good idea to wait for 10.9.1 of Mavericks, as I am sure that will be more stable.

Regards,

Dan Kubley

[charlesludwick · Accepted solution]

@donmontalvo are you seeing machines actually not checking in or running policies? It seems as most of our 10.9 machines work normally, but I am in the middle of confirming it. Just curious to see what others are actually experiencing because of this.

[donmontalvo]

@boettchs Sounds like a Desktop Analyst training issue. :) Point the techs to this thread?

[scottb]

@donmontalvo: It is, sorta. In our environment, there's a bit of "buck-passing" and this is the sort of thing that yes, I do send them info on why it's not the issue, but when these guys and gals want to look good, they scream it's a problem they can't fix because of "Casper". It's just one headache I'd like to go away. But in the end, it's just that - edumacation. Until it's fixed :)

[were_wulff · Accepted solution]

@boettchs

This was fixed in 9.4, though 9.4 introduced another issue with the jamfAgent eating up 40-60% CPU all the time, so I can't say I'd recommend going to 9.4 for that reason alone. That high CPU usage issue was fixed in 9.5, and should still be fixed in 9.51, however, 9.51 has some pretty significant issues with MySQL 5.5 (meaning 'the JSS will crash every couple of minutes') so in the event that you're using MySQL 5.5 in your environment we'd want to either:

1) Upgrade to MySQL 5.6 prior to upgrading to 9.51. The issue is specific to MySQL 5.5.

2) Wait for 9.52, as the defect with MySQL 5.5 should be taken care of in that release. We don't have a concrete release date for 9.52, but I do know the hope was that it would be out sometime this week, though I obviously can't guarantee that will be the case.

If you're still seeing the jamfAgent not responding in 9.5 or 9.51, please get in touch with your Technical Account Manager ASAP to let them know so we can double check and make sure the devices have the correct jamf binary version.

Thanks!
Amanda Wulff
JAMF Software Support

[donmontalvo]

@amamda.wulff thanks for this nugget, we are at 9.32 and holding tight until 9.52 is out and vetted by the community. :)

2) Wait for 9.52, as the defect with MySQL 5.5 should be taken care of in that release. We don't have a concrete release date for 9.52, but I do know the hope was that it would be out sometime this week, though I obviously can't guarantee that will be the case.

[scottb]

@amanda.wulff: Great - thank you. That's good to know, and also thanks for a heads-up on the MySQL stuff. Lord knows I don't want to repeat the 8.73 > 9.32 issues we had. We will certainly wait for at least 9.5.2 then.
Would you recommend to keep MySQL at 5.5 or move to 5.6? I want to try and plan as well as possible for the next upgrade.

[donmontalvo]

@amanda.wulff We hope you're ready for the flood (literally!) of beer that will be waiting for you at JNUC 2014. :D

[were_wulff]

@boettchs

The issue with MySQL 5.5 was completely unintended (and was filed as a defect), and we are still planning to fully support MySQL 5.5 so there shouldn't really be a reason that you'd have to upgrade to MySQL 5.6 unless it was just something you wanted to do to get your environment up to a slightly newer version.

In terms of MySQL 5.7, since we're on the topic of versions, we haven't yet fully tested that one, so I can't say I'd recommend that in a production environment just yet. We do have a few people who are testing it out in dev environments. I haven't had super good luck with it, but others have had an okay time. Still, not something for production just yet.

Amanda Wulff
JAMF Software Support

[were_wulff]

@donmontalvo

Alas, we only get day trips to the JNUC this year (and I work in the Eau Claire office as opposed to the Minneapolis office), so I'll have to leave when our shuttle leaves. :(
Not quite sure what time that is, as the shuttle/ride schedule hasn't yet been finalized.

I'm on the schedule to be there on Tuesday, Sept. 14th, though.

Amanda Wulff
JAMF Software Support

[scottb]

@amanda.wulff - thank you for that. If we don't need to move up, then we'll leave as-is. Your help is much appreciated.

[isaac]

I can report that this is happening in Yosemite (10.10)

[syunetsy]

I can report 10.9.5, and 10.10.2 with JSS 8.73 issue is occurring.

[were_wulff]

Hi @syunetsy

This is a known issue with the 8.x series and any JSS series earlier than 9.4.

In all cases, the solution will be to upgrade to at least 9.6.

In addition, the 8.x series of the JSS does not have 10.10 (or iOS 8) compatibility, so we can’t guarantee you won’t see unexpected behavior when having 10.10 clients enrolled into an 8.x series JSS.
The solution in that case is also to upgrade to 9.6x.

If you have any additional questions or concerns, please feel free to reach out to your Technical Account Manager by giving them a call, sending an e-mail in to support@jamfsoftware.com, or by using the My Support section of JAMF Nation.

Thanks!
Amanda Wulff
JAMF Software Support

[syunetsy]

Thanks amanda.wulff We're in the process of migrating everyone to our new 9.64 JSS instance. :)

[NightFlight]

Currently investigating, but in our 9.93 quickadd package appears to miss putting the SUID and world execution bits on /usr/local/jamf/bin/jamfAgent

Results in a borked Self Service.app

Also randomly misses putting on execution bit on /usr/local/jamf/bin/jamf

issuing /usr/local/jamf/bin/jamf manage fixes the issue. Of course - you might have to add the execution bit manually first.

Results in a really borked casper deployment.

Haven't tested with Recon, but its as stable as it ever was (snicker that's sarcasm BTW).