Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

evacristina

macrumors newbie
Original poster
Sep 8, 2010
26
17
Scenario:
Five brand new late 2013 iMacs bought days apart running 10.9.4 used in an office environment running identical clones of Windows 7 virtually using Parallels Desktop 9.

Symptom:
Only three of the five iMacs 100% randomly and intermittently stop responding (only mouse cursor moves and replies to pings, cannot even SSH, have to force power off) at any given moment without warning (while being used or not) and cannot be replicated and there is nothing in the console logs or Parallels detailed logs to indicate an error or impending trouble. The other two iMacs have been running continuosly for 3 months without a hiccup with the same clone of Windows 7 and same versions of OS X and Parallels.

Troubleshooting performed (assumes much basic troubleshooting in between, i.e. SMC reset, PRAM, permissions, etc):
1) One pass zeroed out disk completely and installed 10.9.4 freshly downloaded from App Store and fresh download of latest build of Parallels 9 - Fail.
2) One pass zeroed out disk completely and installed 10.9.4 freshly downloaded from App Store and fresh download of latest build of VMWARE Fusion 6 - Fail.
3) One pass zeroed out disk completely and installed 10.8.5 freshly downloaded from App Store and fresh download of latest build of Parallels 8 - Fail.
4) Have run Linc's famous script to check for trouble but only shows the usual "Parallels caught causing too many wakes..." which are also present on the working iMacs.
5) Tested memory with Testmem from latest Parted Magic live CD
6) Ran built-in Apple hardware diagnostic

Notes:
* These are plain vanilla installs of nothing except OS X and Parallels.
* The only hardware difference between the working and non-working iMacs is the hard disk;
----- 2 iMacs with APPLE HDD HTS541010A9E662 = Working perfectly
----- 3 iMacs with APPLE HDD ST1000LM024 = 100% randomly stop responding and have to force power off
* All iMacs are set to never sleep, never turn off HDD, never turn off monitor, etc. and are scheduled to shut down every night and startup every morning.
* All iMacs have latest and same version SMC and BootROMs
* I've also tried caffeinating the offending iMacs the whole day to no avail.

Anyone have any insight (Seagate HDD firmware issue?)

-Frank
ACMT, ACSP
 
Last edited:
6) Ran built-in Apple hardware diagnostic
This is kind of vague. Just Apple Diagnostic or ASD/MRI? Did (or can) you run an MRI full surface scan of the disk?

Anyone have any insight (Seagate HDD firmware issue?)
Seagates have a very aggressive (and highly annoying) advanced power management setting that causes the heads to constantly park which causes all sorts of issues. In windows/linux I've used hdparm to disable this behavior. Performance was measurably better. I've never had to do this in OSX so I'm not sure what the command would be. I'm not saying this *is* the problem, but just noting it as a possibility.
 
Have you tried bypassing the internal drive by using an external one?
 
Have you tried bypassing the internal drive by using an external one?

I have not. It may be worth a try, if nothing else to prove to Apple that it's a hardware issue.

Thanks. I'll report back.
 
You mention you've tested the RAM, but have you added any third-party memory to these machines or are they running on stock Apple RAM only?
 
You mention you've tested the RAM, but have you added any third-party memory to these machines or are they running on stock Apple RAM only?

These 5 iMacs are fresh out of the box as plain vanilla installs as they get!
 
These 5 iMacs are fresh out of the box as plain vanilla installs as they get!

Sorry, you didn't specify but I see now by the HDD model numbers that you have the 21.5" iMacs, thus no upgradeable RAM.

What sort of USB devices, if any, are connected to these machines? Are you using them in the VM?

Do these freezes occur only when running Parallels or in OS X as well?
 
Sorry, you didn't specify but I see now by the HDD model numbers that you have the 21.5" iMacs, thus no upgradeable RAM.

Wow! I can't believe I ommitted 21.5" Sorry!

What sort of USB devices, if any, are connected to these machines? Are you using them in the VM?

One USB signature pad configured to connect to the VM every time. The offending iMacs froze just the same without any USB devices connected. To add insult, the behaving iMacs' USB ports have the same USB signature pads plus an x-ray sensor plugged in.

Do these freezes occur only when running Parallels or in OS X as well?

Both the behaving and offending iMacs run Parallels full screen all day. Windows shuts down automatically every night but Parallels keeps running full screen so they never see Mac OS X.

I'll be running memtest86 again tonight but in parallel CPU mode. Then, I'm taking my case to the geniuses tomorrow morning. I'll keep updating, hopefully to save somebody this hasle in the future.
 
*** Update

After some time spent at the genius bar, they've agreed to replace the Samsung hard drive with a Hitachi part. I will post the findings after some long term testing.
 
Scenario:
Five brand new late 2013 iMacs bought days apart running 10.9.4 used in an office environment running identical clones of Windows 7 virtually using Parallels Desktop 9.

Symptom:
Only three of the five iMacs 100% randomly and intermittently stop responding (only mouse cursor moves and replies to pings, cannot even SSH, have to force power off) at any given moment without warning (while being used or not) and cannot be replicated and there is nothing in the console logs or Parallels detailed logs to indicate an error or impending trouble. The other two iMacs have been running continuosly for 3 months without a hiccup with the same clone of Windows 7 and same versions of OS X and Parallels.

Troubleshooting performed (assumes much basic troubleshooting in between, i.e. SMC reset, PRAM, permissions, etc):
1) One pass zeroed out disk completely and installed 10.9.4 freshly downloaded from App Store and fresh download of latest build of Parallels 9 - Fail.
2) One pass zeroed out disk completely and installed 10.9.4 freshly downloaded from App Store and fresh download of latest build of VMWARE Fusion 6 - Fail.
3) One pass zeroed out disk completely and installed 10.8.5 freshly downloaded from App Store and fresh download of latest build of Parallels 8 - Fail.
4) Have run Linc's famous script to check for trouble but only shows the usual "Parallels caught causing too many wakes..." which are also present on the working iMacs.
5) Tested memory with Testmem from latest Parted Magic live CD
6) Ran built-in Apple hardware diagnostic

Notes:
* These are plain vanilla installs of nothing except OS X and Parallels.
* The only hardware difference between the working and non-working iMacs is the hard disk;
----- 2 iMacs with APPLE HDD HTS541010A9E662 = Working perfectly
----- 3 iMacs with APPLE HDD ST1000LM024 = 100% randomly stop responding and have to force power off
* All iMacs are set to never sleep, never turn off HDD, never turn off monitor, etc. and are scheduled to shut down every night and startup every morning.
* All iMacs have latest and same version SMC and BootROMs
* I've also tried caffeinating the offending iMacs the whole day to no avail.

Anyone have any insight (Seagate HDD firmware issue?)

-Frank
ACMT, ACSP

Do you get it to fail using anything other than Parallels or Fusion? Contrastingly, is there any chance that your Windows 7 VM image is corrupted?

These are questions that any ACMTed technician or Genius Bar goon would ask you prior to checking in any of those machines for service.

Also, before we label anything up with the firmware of the hard drive common to those iMacs, have you tried the exact same software environment on a different kind of Mac with a different drive entirely?
 
Do you get it to fail using anything other than Parallels or Fusion? Contrastingly, is there any chance that your Windows 7 VM image is corrupted?

It failed under Parallels 9, Parallels 8, VMWare Fusion 6. I've copied the VM image many times in different ways from the known working on the two other iMacs.

Also, before we label anything up with the firmware of the hard drive common to those iMacs,

It turned out to be a Samsung part and not Seagate. This Samsung part has a dismal reliability record.

have you tried the exact same software environment on a different kind of Mac with a different drive entirely?

Yes, the same VM image is working flawlessly on a 2011 MacBook Air with 10.9.4 and Parallels 9 in the same office.

By the way, I've been troubleshooting this for 3 months now and have spent about 100 hours through a process of elimination. I've even taken into to consideration the possibility of RFI interfering with the operation of the drive. I don't know if these Samsung parts have Spread Spectrum enabled by default or not.

I have not received the repaired iMac yet which is strange. I hope they sent it to the real geniuses for testing.
 
9-24-14 Update:

The iMac's Samsung ST1000LM024 hard drive was replaced by the Apple store techs with the same APPLE HDD HTS541010A9E662 that shipped on the working iMacs and 10 days later - not one freeze. I'll be taking the second iMac with the Samsung part in to the Apple shop to be replaced as well. I hope this helps anyone out there struggling with random and intermittent freezes with this Samsung ST1000LM024 part in their Macs.
 
11-10-14 Update:

After three weeks of the engineers at Apple trying to figure out what the problem might be, the second of 3 iMac's Samsung ST1000LM024 hard drive was replaced by the Apple store techs with the same APPLE HDD HTS541010A9E662 that shipped on the working iMacs and 10 days later - not one freeze. I'll be taking the third iMac with the Samsung part in to the Apple shop to be replaced as well. I hope this helps anyone out there struggling with random and intermittent freezes with this Samsung ST1000LM024 part in their Macs. And I hope this also helps the skeptics out there still wondering what the problem might be :)
 
Thanks for posting! I dropped off a 27-in iMac this morning for identical symptoms -- stops responding except for the mouse cursor, responsive to pings but not ssh. Were you able to identify the drives as Samsung before they were removed? I have a different drive model but wonder if it is Samsung or Seagate. The drive model is "APPLE HDD ST1000DM003".

In trying to diagnose what I could, I had booted to a usb recovery partition and run Disk Utility multiple times. Eventually, while checking either the disk or permissions, it would error out despite having run successfully prior to that. This machine has the fusion drive, and after the error occurs the drive turns red in Disk Utility's list (see attached). Clicking on the drive shows the message "This Fusion Drive is missing a disk." Checking from terminal the device is indeed no longer present (/dev/disk1 in my case), along with the combined fusion drive (disk3). The associated SSD drive is still present. Also, having started Disk Utility from the command-line it produced a long list of "Operation not permitted" and "Device not configured" errors. A cold boot is necessary for the drive to show up again.

Checking the drive's SMART error log shows 0 errors, and running the extended test ends successfully.
 

Attachments

  • screenshot-error.png
    screenshot-error.png
    144.7 KB · Views: 276
XtrEm3:

Yes, the drive was confirmed to be a Samsung part by the Genius Bar techs. At first it was confusing because I believe Seagate bough Samsung's drive business or the other way around. The next time I buy a Mac, if it comes with a Samsung mechanical drive, I will return it.
 
I agree.

The OP could've just gone for a pure-SSD iMac straight away and avoid any problems with spinners :rolleyes:

I agree completely, especially considering how slow Yosemite would run on 5400 RPM rust...
 
* These are plain vanilla installs of nothing except OS X and Parallels.
* The only hardware difference between the working and non-working iMacs is the hard disk;
----- 2 iMacs with APPLE HDD HTS541010A9E662 = Working perfectly
----- 3 iMacs with APPLE HDD ST1000LM024 = 100% randomly stop responding and have to force power off

Just googled:

HTS541010A9E662 = Hitachi Travelstar

http://www.hgst.com/tech/techlib.nsf/techdocs/79BA810F702AD64E88257977006375D8/$file/TS5K1000_ds.pdf

ST1000LM024 = Seagate Momentus

http://www.seagate.com/files/staticfiles/support/docs/samsung-ds/100698122c.pdf

Internet gosspis says, that is beacause Seagate abandoned quality control department.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.