I have had two Mac Mini workstations (10.10.x) in a small office give me weird behavior. The server is a 10.10.4-based Mac Mini. All networking is done over ethernet.

I thought I had solved it on the first computer with some network infrastructure fixes and a re-imaging of the workstation, but the problem appeared on a second workstation months later. A network user logs in, and gets an error message of: “The home folder for user ‘xxxx’ isn’t located in the usual place or can’t be accessed.” When the user logs into another workstation, everything works fine.

Checking from terminal (df), their home folder appears to be mapped but their Desktop contents are not displayed on the screen. If I list files in the Desktop in terminal, all the files are there. Looking at Finder’s Go list of file locations, “Home” is grayed out. These signs make me think the problem is with Finder.

With Google I’ve found a couple of references to similar problems, the closest match being five years ago. The person solved the problem by deleting the extended attribute com.apple.FinderInfo (sudo xattr -d com.apple.FinderInfo /). However, when I try xattr -P, I get a response that the attribute does not exist. Not sure if Apple changed this or the solution simply isn’t relevant in my case. I also found mentions of Kerberos, but since this isn’t user specific I’m not sure that would be germane. Thoughts from any Kerberos experts?

I’ve un-enrolled the workstation and deleted all the profiles. Re-enrolled. Also rebound to the OD. I may be left in a situation of re-imaging this workstation too, but I’m hoping someone else has seen this or knows what I should try to fix?

Thank you.


| improve this question | | | | |
  • Did you follow all steps in Linc Davis` reply to this "question": "The home folder for user X isn't located in the usual place or can't be accessed."? – klanomath Aug 21 '15 at 21:14
  • With the exception of steps 4 and 9, yes. Because only one workstation at a time is affected, I'll definitely skip #9 for now. I'll research #4 more. I am doing another server reboot tonight after the clients are off. – tim.rohrer Aug 21 '15 at 22:46
  • I read the link in Step 4 and I have two questions: 1) Does anyone know how important the note is about listing the server as in DNS? I don't think I've ever done that. I'm pretty sure it is listed with the static ip I gave it. 2) What impact with rebuilding Kerberos config have on users who are not having troubles on other workstations? This step is starting to make sense to me, but it worries me about making matters worse. Thanks. – tim.rohrer Aug 21 '15 at 22:53
  • Did you ever find a solution? I have the same situation; only a handful of machines are affected, but it seems to be spreading. Re-imaging does fix it for single affected workstations, but I'd prefer to prevent the problem... – Andreas Ley Mar 3 '16 at 18:24
  • No. And I ended up leaving the client where it was happening. I've not seen it with other deployments. – tim.rohrer Mar 19 '16 at 14:49

You must log in to answer this question.

Browse other questions tagged .