Versions working erratically, with xpchelper errors

I will go for the "me too" as well.

The problem occurred while casually switching from my internet browser to Microsoft Word on a non-admin account.

The beach ball appeared, my fans went berserk, Word was flagged as "Not Responding".

I attempted to launch the Activity Monitor to find out what was holding up my computer resources.

Unfortunately, the application never properly launched and the fans kept on screaming.

I closed and/or force-quitted all applications with no effect on my CPU.

I tried to open the Terminal but the OS displayed an error message going along the lines of "the administrator set an illegal value to your shell" and quitted.

I switched to the admin user account and stumbled upon xpchelper persistently spamming the logs as described in this thread.

I found the process appearing/disappearing on the Activity Monitor and managed to shut it down from the admin account.

Everything went back to normal... (except that I've lost access to my Apple ID account (probably unrelated problem) and ended up using my gf's to post here)

In 10.7.3 the xpchelper executable has been recompiled (27th Jan 2012) - Hopefully this means that this bug has been crushed. 🙂

updated to 10.7.3 yesterday and the problem again just now.

Happened to me twice for now.

First case:

System: Upgraded to 10.7.3 from snow leopard.

Months later it went nuts.

ssh yield You don't exist, go away!

whoami yield 501 instead of "noice" (my username).

Terminal.app refused to start - no shell for user 501.

iTerm started, but recognized me as 501 instead of noice as I said above.

Console was full of this message as you do.

I started to panic, nothing helped. I didn't want to reboot as I was afraid I will not be able to log in - if the system doesn't know me.

/etc/passwd doesn't contain "noice" at all, it referes to opendirectory daemon.

After some time I did clean install on new SSD. No timemachine restore or something like that. Just pure install with fresh installed apps.

Now, when I just needed to connect to some server via ssh and I saw that "You don't exists, go away!" message AGAIN. "whoami" reports 501 as usual instead of "noice".

So I checked the Console - it was there all like before.

It went away and started "to know me" after 20 minutes or so.

So I still don't know what causes this, I was doing the same things as always.

The last things I did:

- connected external disk (for timemachine) cca 10 minutes before it happened

- connected to some server via RDP

- had 8 terminals in iTerm 2 open, 3 of which were connected to remote hosts

Open apps: adium, itunes, sourcetree, sequel pro, evernote, sublime text 2, chrome, iterm 2, apple mail.

I also noticed that the biggest resource hogs were kernel_task (I think that's due to logging), Mail (50% cpu, but that can be short-lived coincidence...).

I recently read that killing the opendirectoy daemon works as a workaround - I'll try it next time.

I tried that the first time.. problem is.. that you cannot kill it with sudo kill -9 opendirectoryd ... because you can't use sudo when the system doesn't know you. "You are not in sudoers file."

Activity monitor can't kill root processess too...

Right, but I'm working as a standard user and I still can switch to the administrator account, which works normally 🙂

via su or sudo?

sudo doesn't work... or you would have to be sudo-ed before it happens...

su doesn't work for me (as I am lazy to type sudo su and change the root password)

NoICEReaver, have you tried the solution via Fast user switching (it's described earlier on the thread by user acrollet)? It helped at least two people.

No, I don't have another account on this mac. But I will try it next time it happens...

Add me as a "me too". Just had this happen under 10.7.3 last night for the first time. Spotted it when I got the same "You do not exist" messages when trying to ssh into some work systems. Cleared itself up after an hour or so. Can't see anything in the console logs that hints out how it started or how it cleared up.

Another me too.

It happened to me today just after wake on 10.7.3. The error just kept spamming the console and I had no rights to open the terminal as well. After the reboot everything seems to be normal again.

Just another parrot here - me too, me too. Mine occurred for the first time today (10.7.3) and I wasn't even at the computer when it manifested itself. I was making lunch and I heard the fans on the Macbook whirl up. Went to see, tried to launch activity monitor and it wouldn't - bounce, bounce, bounce.

Found all the errors (coming too fast and furious to even scroll back and read easily), but couldn't isolate a culprit. In my case some combination of force quitting the finder and fast user switching out and back into my (admin) account put a stop to the errors. It sort of petered out with this series of messages in the console:

...pages of the same error, then...

Apr 6 13:54:33 ********** xpchelper[74898]: getpwuid_r() failed for UID: 501, ret: 0, errno: 0

Apr 6 13:54:33 ********** com.apple.launchd[1] (com.apple.xpchelper.F5010000-0000-0000-0000-000000000000[74898]): Exited with code: 1

Apr 6 13:54:33 ********** xpchelper[74899]: getpwuid_r() failed for UID: 501, ret: 0, errno: 0

Apr 6 13:54:33 ********** com.apple.launchd[1] (com.apple.xpchelper.F5010000-0000-0000-0000-000000000000[74899]): Exited with code: 1

Apr 6 13:54:33 ********** xpchelper[74900]: getpwuid_r() failed for UID: 501, ret: 0, errno: 0

Apr 6 13:54:33 ********** com.apple.launchd[1] (com.apple.xpchelper.F5010000-0000-0000-0000-000000000000[74900]): Exited with code: 1

Apr 6 13:54:33 ********** xpchelper[74901]: getpwuid_r() failed for UID: 501, ret: 0, errno: 0

Apr 6 13:54:33 ********** com.apple.launchd[1] (com.apple.xpchelper.F5010000-0000-0000-0000-000000000000[74901]): Exited with code: 1

Apr 6 13:54:34 ********** System Preferences[74366]: Can't ODFDEResetPassword because of NULL unlockPassword!

Apr 6 13:54:38 ********** iCal[54570]: ServerNotifications: Setting delegate to APSD

Apr 6 13:54:40 ********** webdavfs_agent[12298]: WebDAV server is now responding normally

Apr 6 13:54:40 ********** KernelEventAgent[66]: tid 00000000 received event(s) VQ_NOTRESP (1)

Apr 6 13:54:43 ********** SecurityAgent[74906]: Echo enabled

Apr 6 13:54:44 ********** iTunes[54940]: The plugin 'Quartz Composer Visualizer' failed to load because it has the wrong CPU architecture for this version of iTunes. (3585)

Apr 6 13:54:46 ********** iCal[54570]: Subscription request completed

Apr 6 13:54:46 ********** SecurityAgent[74906]: User info context values set for testuser

Apr 6 13:54:47 ********** SecurityAgent[74906]: Login Window login proceeding

Just for completeness, "testuser" is the userid I fast-switched to from my admin account. Messages were more or less normal after that.

-c

Me too.

After 73 days uptime I just got the bug again, but finally found out how to fix it (or: workaround):

I'm working as a normal user (non-Admin).

- The bug occured and I used fast user switching, to switch into the admin account.

- There I opened up the Activity Monitor and simply quit the opendirectoryd (automatically restarts itself).

- CPU usage went back to normal and my user account is fully responsive again, without having to restart.

Hopefully this helps you next time 🙂