The Hard Drive Click Of Death: Two Weeks, Two Mortalities?

Wednesday was a particularly great day…at least at first. Down in Silicon Valley for an Adobe briefing on Monday afternoon and EDN meetings all day Tuesday, I had breakfast with Analog Devices yesterday morning. Then, before heading for home, I swung over to VMware for a briefing on, and guided upgrade to, version 2 of the company’s Fusion virtualization product.

Before leaving VMware, I closed the MacBook’s bezel (therefore putting the system to sleep) with upgraded Fusion still among the active programs. Since Fusion (and Windows XP Pro SP3 running on it) appears to OS 10.4 as just another application, this approach worked fine with Fusion v1, leveraging OS X’s generally rock-solid power management scheme. When I got back to Truckee last night, I woke the MacBook back up…and was greeted by the following ominous SMARTReporter pop-up:

OS X’s Disk Utility echoed the reported hard drive degradation:

And although neither of the above utilities gave me the exact S.M.A.R.T. error code, SMART Utility came through with the data I was seeking:

Frustrating? Darn right it was, especially considering I’d just put this new hard drive into the system less than two weeks ago. I immediately backed up my Outlook database to external storage then, seeing that the hard drive still seemed to be alive, I also archived my entire Fusion virtual machine image. The hard drive still seemed to have a strong pulse, so I then backed up the entire HDD using SuperDuper! Then I went to bed.

When I woke up the MacBook this morning, I again saw pop-up evidence of S.M.A.R.T. spin-up woes but in mentally revisiting the situation, I discerned that it was mighty coincidental that the problem emergence coincided with my Fusion v1.1.3-to-2 update. Also, take a close look at the above SMART Utility screenshots. They suggest a nearly 23 second HDD spin-up latency (the unit of measure is msec); although I didn’t put a stopwatch on the system wake-up cycle, there’s no way it was that long in reality.

On a hunch, I suspended the Windows virtual machine and quit Fusion, then put the MacBook back to sleep. When I re-awoke the computer an hour later, the S.M.A.R.T. monitoring utilities reported no re-emergence of the error. Unfortunately, since S.M.A.R.T. data cannot be cleared once set, for perhaps obvious reasons—otherwise you’d have folks selling bad drives on Ebay, for example—this HDD will carry the evidence of past spin-up woes for the rest of its days. But based on my limited experimentation to date, I’m leaning towards the theory that the hard drive is actually fine and Fusion v2 is somehow the root cause of the problem I’m seeing.

I’ve forwarded this information to my VMware contact and I’ll report back anything the company discovers. While there’s a lot to like about the v2 Fusion upgrade (elaboration on which I’ll save for a dedicated post to come soon), if my theory’s right, this is one ugly bug that VMware needs to squash…and soon.

p.s….for comparison’s sake, by the way, here’s a screenshot of the SMART Utility report after the system’s awoken without Fusion launched:

Keep in mind in comparing the two screenshots that ‘value’, ‘worst’ and ‘threshold’ are normalized numbers, not absolutes, and that ‘bigger is better’ with respect to ‘value’.