The application Finder.app can't be opened.

Wilson, thanks - I thought there must be a way but couldn't remember what it was. Mine seems to be working fine now, but I might try the complete re-index if it recurs.

ph.

I tried a simple stop & restart – that is, "sudo mdutil -i off /" and then "sudo mdutil -i on /" – but it still seems to take forever, and logs show this sort of error:

Thread 2 Crashed:
0 ...ple.CoreServices.CarbonCore 0x00007fff8174be7f CSStoreGetUnit + 84
1 com.apple.LaunchServices 0x00007fff8725123f _LSContainerCheckState + 65
2 com.apple.LaunchServices 0x00007fff8726806c _LSCopyLibraryItemURLs + 419
3 mdworker 0x0000000100003795 0x100000000 + 14229
4 mdworker 0x00000001000040b2 0x100000000 + 16562
5 mdworker 0x0000000100004583 0x100000000 + 17795
6 mdworker 0x0000000100008f32 0x100000000 + 36658
7 libSystem.B.dylib 0x00007fff806588b6 pthreadstart + 331
8 libSystem.B.dylib 0x00007fff80658769 thread_start + 13

...which I assume means my indexing is going to go nowhere.

So this time, I tried the above two commands again, but with "sudo mdutil -E /" sandwiched in between. I'd have to give it hours to see whether anything is going to happen (still just "Estimating indexing time"), but unfortunately, Console reveals the same sort of error:

Thread 2 Crashed:
0 ...ple.CoreServices.CarbonCore 0x00007fff8174be7f CSStoreGetUnit + 84
1 com.apple.LaunchServices 0x00007fff8725123f _LSContainerCheckState + 65
2 com.apple.LaunchServices 0x00007fff8726806c _LSCopyLibraryItemURLs + 419
3 mdworker 0x0000000100003795 0x100000000 + 14229
4 mdworker 0x00000001000040b2 0x100000000 + 16562
5 mdworker 0x0000000100004583 0x100000000 + 17795
6 mdworker 0x0000000100008f32 0x100000000 + 36658
7 libSystem.B.dylib 0x00007fff806588b6 pthreadstart + 331
8 libSystem.B.dylib 0x00007fff80658769 thread_start + 13

Is there something else I should be doing in order to perform a "complete refresh"?

(In any case, since 10.6.3 I'm still not seeing any return of the system malfunctions that previously had been triggered by indexing, so that's good. I just can't get Spotlight indexing and working again!)

I'm guessing the "sudo mdutil -E /" would erase the indexes so they'd be recreated next time.

You could check if the index files have gone. I don't remember if they're stored on the drive itself or in a common place (eg /var somewhere).

Looks like it's mdworker crashing in your case from those stack traces, and that does the spotlight indexing - http://developer.apple.com/mac/library/documentation/Darwin/Reference/ManPages/m an8/mdworker.8.html

This page might be useful for you:

http://www.thexlab.com/faqs/stopspotlightindex.html

ph.

you could also try using OnyX to rebuild the spotlight databases:

http://www.macupdate.com/info.php/id/11582/onyx

though it may just do what you've already done. It may also be able to clear any preferences files and caches for spotlight, which may be causing problems.

ph.

Looks like I spoke too soon when I said I'd fixed my issues. There were no issues all day and the logs looked fine. Then I let Software Update install iTunes 9.1 and the latest iPhoto 8.1.2 and the problems recurred. Both seemed to install fine then the restart dialog appeared asking for a password and that's when it started going wrong - keypresses went to the wrong application, menu bar disappeared, couldn't log out, couldn't ping the machine from another machine, couldn't launch apps from the dock. Had to do a hard reboot again.

This is the only stuff in the logs around that time:

31/03/2010 20:10:07 com.apple.audio.coreaudiod[163] coreaudiod: CarbonCore.framework: coreservicesd process died; attempting to reconnect but future use may result in erroneous behavior.
31/03/2010 20:10:07 com.wacom.wacomtablet[370] TabletDriver[370]: CarbonCore.framework: coreservicesd process died; attempting to reconnect but future use may result in erroneous behavior.
31/03/2010 20:10:07 com.apple.launchd[1] (com.apple.coreservicesd[66]) Job appears to have crashed: Segmentation fault
31/03/2010 20:10:08 com.wacom.wacomtablet[370] TabletDriver[370]: CarbonCore.framework: coreservicesd process died; attempting to reconnect but future use may result in erroneous behavior.
31/03/2010 20:10:08 com.apple.ReportCrash.Root[1826] 2010-03-31 20:10:08.680 ReportCrash[1826:2803] Saved crash report for coreservicesd[66] version ??? (???) to /Library/Logs/DiagnosticReports/coreservicesd 2010-03-31-201008localhost.crash
31/03/2010 20:12:43 com.apple.launchd[1] (com.apple.suhelperd[879]) Exited with exit code: 2

A crash of coreservicesd seems to be root issue. So it doesn't look like an issue with mdworker as before, but there's still something not right.

Process: coreservicesd [66]
Path: /System/Library/Frameworks/CoreServices.framework/Versions/A/Frameworks/CarbonC ore.framework/Versions/A/Support/coreservicesd
Identifier: coreservicesd
Version: ??? (???)
Code Type: X86-64 (Native)
Parent Process: launchd [1]

Date/Time: 2010-03-31 20:10:05.889 +0100
OS Version: Mac OS X 10.6.3 (10D573)
Report Version: 6

Exception Type: EXC BADACCESS (SIGSEGV)
Exception Codes: KERN INVALIDADDRESS at 0x00000001079b0064
Crashed Thread: 0 Dispatch queue: com.apple.main-thread

Thread 0 Crashed: Dispatch queue: com.apple.main-thread
0 ...ple.CoreServices.CarbonCore 0x00007fff8583190f _FreeSegGetLength(void const*, unsigned int) + 15
1 ...ple.CoreServices.CarbonCore 0x00007fff8584bd64 _TableAllocBytes(void*, unsigned int, unsigned int*) + 102
2 ...ple.CoreServices.CarbonCore 0x00007fff8584bbed CSStoreAllocBytes(_CSStore*, unsigned int, unsigned int, unsigned int*) + 91
3 ...ple.CoreServices.CarbonCore 0x00007fff8584bec2 CSStoreAllocUnitWithData + 77
4 com.apple.LaunchServices 0x00007fff811f730f CSCopyStringForCharacters + 107
5 com.apple.LaunchServices 0x00007fff811f7276 _LSDatabaseCreateStringForCFString + 98
6 com.apple.LaunchServices 0x00007fff81223cdc _LSServerItemInfoRegistration + 3222
7 com.apple.LaunchServices 0x00007fff8122cdef LSServerRegisterItemInfo + 128
8 com.apple.LaunchServices 0x00007fff81235ae9 LSPluginServerRegisterItemInfo + 86
9 com.apple.LaunchServices 0x00007fff8122e4bc _XRegisterItemInfo + 373
10 com.apple.LaunchServices 0x00007fff8122dbe6 LSPlugin_server + 106
11 ...ple.CoreServices.CarbonCore 0x00007fff858b7a81 scHandleMessage(mach msg_headert*, unsigned int ( )(machmsg_headert, mach msg_headert*), unsigned int*, unsigned char*) + 119
12 ...ple.CoreServices.CarbonCore 0x00007fff858b5328 SCServerSession::handleMessage(mach msg_headert*) + 370
13 ...ple.CoreServices.CarbonCore 0x00007fff858b454c serverCallback(_CFMachPort*, void*, long, void*) + 342
14 com.apple.CoreFoundation 0x00007fff845cdc2e __CFMachPortPerform + 366
15 com.apple.CoreFoundation 0x00007fff845a6201 __CFRunLoopRun + 5201
16 com.apple.CoreFoundation 0x00007fff845a48df CFRunLoopRunSpecific + 575
17 com.apple.CoreFoundation 0x00007fff845a4666 CFRunLoopRun + 70
18 ...ple.CoreServices.CarbonCore 0x00007fff858b42c2 _CoreServicesServerMain + 522
19 coreservicesd 0x0000000100000eac 0x100000000 + 3756

Thread 1: Dispatch queue: com.apple.libdispatch-manager
0 libSystem.B.dylib 0x00007fff834674ea kevent + 10
1 libSystem.B.dylib 0x00007fff834693bd dispatch_mgrinvoke + 154
2 libSystem.B.dylib 0x00007fff83469094 dispatch_queueinvoke + 185
3 libSystem.B.dylib 0x00007fff83468bbe dispatch_workerthread2 + 252
4 libSystem.B.dylib 0x00007fff834684e8 pthreadwqthread + 353
5 libSystem.B.dylib 0x00007fff83468385 start_wqthread + 13

Thread 2:
0 libSystem.B.dylib 0x00007fff83489316 _semwaitsignal + 10
1 libSystem.B.dylib 0x00007fff8348d131 pthread_condwait + 1286
2 ...ple.CoreServices.CarbonCore 0x00007fff8586829a fmodWatchConsumer + 347
3 libSystem.B.dylib 0x00007fff834878b6 pthreadstart + 331
4 libSystem.B.dylib 0x00007fff83487769 thread_start + 13

Thread 3:
0 libSystem.B.dylib 0x00007fff83458d46 read + 10
1 ...ple.CoreServices.CarbonCore 0x00007fff85868ff5 fmodWatchProducer + 140
2 libSystem.B.dylib 0x00007fff834878b6 pthreadstart + 331
3 libSystem.B.dylib 0x00007fff83487769 thread_start + 13

Thread 0 crashed with X86 Thread State (64-bit):
rax: 0x0000000000760064 rbx: 0x00000001079b0064 rcx: 0x00007fff5fbfd76c rdx: 0x000000010734f8dc
rdi: 0x0000000107250000 rsi: 0x0000000000760064 rbp: 0x00007fff5fbfd680 rsp: 0x00007fff5fbfd670
r8: 0x00007fff5fbfd7bc r9: 0x00007fff5fbfd77c r10: 0x00000001001e29c0 r11: 0x00007fff5fbfd81c
r12: 0x0000000107250000 r13: 0x00000000000ff8dc r14: 0x000000000000001c r15: 0x000000000000073c
rip: 0x00007fff8583190f rfl: 0x0000000000010202 cr2: 0x00000001079b0064

Like everyone else here... "me too." I just installed 10.6.3 and my beautiful Mac started giving me the headaches I only get with Windows!

I have something that might help:
In between lockups, I managed to get the Console Messages (HD/Utilities/Console) up. I work in Unix/Linux most of the time, so logs and bash are my happy places. The logs revealed that I had a failing tablet driver:

I/O
warning:
failed to load external entity "/Library/Preferences/
com.pentablet.defaults.xml

Luckily, Wacom provided a way to easily remove the driver (go to the folder in Applications and use the Uninstall option - thanks, Wacom!)

I also rebuilt the LaunchService database from Terminal (this is one line, without the \), though I don't claim that this helped or hurt:
/System/Library/Frameworks/ApplicationServices.framework/\
Frameworks/LaunchServices.framework/Support/lsregister \
-kill -r -domain local -domain system -domain user

http://www.macosxhints.com/article.php?story=20031215144430486

So, go look for something failing ("warnings"). I've read a lot on this subject and it seems that Time Machine and Disk Utility are symptoms - its what they access that is causing the problem (NFS shares, SMB shares, USB drives, external drives), because of a driver or connection made out-of-date by an OS update.

I guess the rule of thumb for OS X users should be: Before you update, go check for new drivers for non-Apple hardware. If you can't find any, remove the old hardware and any drivers it came with, then do the update. If there is an NFS or SMB share causing problems, make sure the host OS has been properly patches.

Phil,

I suspect you may have something physical going down. All your symptoms sound a bit like my Mk1 Intel iMac (the 2006 one). I think Apple replaced everything apart from the screen and the case before giving up and replacing the entire machine with a Core2 Duo one. As a start, you could run Tech Tools.

Wilson

Inked,

Your discovery on the Wacom is very interesting. I bought and installed an Intuos 4 just before Snow Leopard on my iMac. My iMac was by far the worst of my Macs for the 10810 problem. The Wacom is currently sitting on my PowerMac working well with Leopard. When I go back down to France in two weeks and the Wacom goes back onto my iMac, I will take your advice and uninstall the current driver and reinstall the latest. I was using a Snow Leopard driver for the Wacom but it sounds as if it was not good enough.

Wilson

It's possible it's a hardware fault. If it is, hopefully it's just RAM. I'll try some ram tests and techtool etc.

Some certainly seems to be a software issue though. Most of the crashes were happening with mdworker, which has gone away after I disabled and re-enabled the NFS mounts, seemingly forcing spotlight to drop and re-create the indexes for the drives. Before that I was getting 20+ crashes a day. After that I've had 1 in 36 hours.

I think it all started when I was adding stuff to my music collection (which is on an NFS mount) via iTunes. That was importing some music while another program was re-encoding some FLAC files to ALAC on the same NFS mount. The system hung for some reason, and I think it's due to Spotlight and what I was doing with the NFS drives at the time.

Before then I'd never seen this crash. After that I was getting it many times per day.

I've also seen a couple of crashes that seemed to be related to a Wacom Tablet driver, which has gone away since I re-installed the latest drivers. Though I'm still seeing errors in the logs for it. Perhaps I'll send those to Wacom for inspection.

The latest crash was from coreservicesd and happened as I was installing updates to iTunes and iPhoto using Software Update. The stack trace didn't give me any idea why it crashed though.

I'll certainly check for hardware issues anyway, so thanks for that suggestion.

Huzzah. Partial relief was obtained when I deleted (actually just moved out of the way) the contents of /var/folders. I am now able to complete an index build with Spotlight including the /Applications and /Library directories, which I was not before. /var/folders holds temporary data for a number of programs, and apparently spotlight keeps some portion of an sqlite database there. I had assumed that 'mdutil -E /' would actually empty that out but I guess some corrupted pieces stayed behind. I deleted everything in /var/folders but you could trace the mds/mdworker processes to see exactly which dir spotlight is building its db in on your machine.

Anyway, to those folks whose coreservicesd was crashing due to Spotlight/mdworker let me know how it goes with you, or if you've already been down this path. I don't believe this is a permanent solution since whatever caused the corruption is still out there.

Apple has ****** me off with this tremendously, but at least they gave me dtrace so I can try to puzzle it out on my own. Now I just need them to give me the source code so I can fix it 🙂

Like many on this thread, I have been bitterly disappointed and disillusioned by Apple's desperately weak response to this issue. As soon as I turn indexing back on for an External Storage device, the problem re-appears. For those who would like more control over Spotlight's behaviour, may I suggest Spotless 3 shareware utility. I suspect I will be buying a copy in the next week or so.

Wilson

For those looking for a source of your troubles - go into Applications/Utilities/Console.

Don't look for events that happen at the time of the crash. Look for events after the last reboot, prior to coreservicesd getting all Microsofty and throwing crashes up in your grill (sorry - had to throw that in for the "younger crowd" I probably already lost them by using grammar and spelling my words out).

Then, go line-by-line.

"Warnings," are a good place to start. Right now, VMware is giving me a ton of warnings, but they're ok. Now, warnings for devices where, "XML file can't be read," those are bad. What does that mean? Well, not having coreservicesd.c in front of me, I can only guess that means, "I tried to read it, but it didn't have the stuff I was expecting, so, while I'm going to make it available (source of the crash), it probably won't work so well."

I'll bet I could go back and re-install that bad Wacom driver and OS X 10.6.3 would adopt it and make it its own (if it finished the install, that is). Chances are, 10.6.2 created that XML file in the first place and that's why 10.6.3 had so many problems. That explains the error messages we're seeing, about memory addresses being corrupt: that's where the missing driver should be. Coreservicesd is trying to serve it up, but can't, so crash, crash, crash.

Think of it this way: you go into the hospital with a cracked, bleeding skull. The doctor patches you up. How'd you get the cracked, bleeding skull? Fell on the concrete. So, you might think, "remove the concrete, or make it soft and pillowy." Then, the doctor releases you and smash, you crack your skull again on the hospital floor. You just keep crashing. The doctor can keep patching you, but you haven't even looked at the source of the problem: your shoes are tied together. Get it? Go back to the beginning, go back to boot and start there, not at the time around the crash, that's pointless.

Don't lose heart, friends. I, too, was preparing a Jobs picture to hang up next to Gates, Hitler and O...., but Steve & Co. (and BSD) at least left us enough clues in the console log to help ourselves. So, Steve goes back on the shelf, next to Reagan, Hello Kitty and my Big Boy bank collection.

Think of the alternative: you don't really want to install... You Know What 7... do you? Shoot. I wouldn't even waste bandwidth trying to torrent it, let alone buy it and convince magnetic spots on my hard drive to store it.

Huzzah redux! I found a solution in the advice of jitterysquid.

To recap my situation:

1) I narrowed down Spotlight indexing as the trigger for my 10810 syndrome. (For lack of a name, that's what I'll call the whole mess of problems described by me and others, often accompanied by Finder -10810 errors.)

2) Upgrading to 10.6.3 appears (so far) to have cured my 10810 syndrome. But Spotlight indexing would stall on either the "Estimating" or the "Indexing" stage, never finishing an index and becoming usable. I tried many Spotlight resetting operations described on this page and elsewhere, to no avail.

3) I tried disabling Spotlight, deleting the content of /var/folders, then reenabling Spotlight. It soon started indexing, displayed a reasonable time estimate of a few hours, and did the job. Everything seems to work perfectly now!

(Also FYI: In my situation the presence or absence of external drives didn't seem to make any difference in things, before or after the fix. That may differ from many others' experience.)

Thanks to everyone commenting with helpful ideas! Is anyone else trying out the /var/folders technique?

Sorry to be such a dunce, but where is /var/folders located? I can't seem to find it so I assume it's buried somewhere in the System folder.

In the Finder look at the "Go" menu: "Go to folder...". You can paste "/var/" in there and it will open that folder as a regular window.