Death of a Hard Drive


It must have been crawling along for weeks if not months, but I never noticed it until last night when the hard drive that holds all of my source code enlistments decided to take a dive for the worst. I’ve put SpinRite on it, but at current pace it is predicting completion in 30+ days and most sectors are coming back Unrecoverable… sigh I don’t think I’m going to stop working for 30 days. And how dare it stop working on my birthday, too… :-/


Fortunately, it is just source code enlistments to IIS5, IIS6, IIS7, and several other projects, which means I am only losing the client-side cached copies of source code along with any of my unsubmitted changes, and the only corruptions I care about are in those textual source files. But, not too many files are in that state, so recovery should not stressful. I just need to get a new hard drive, re-establish and re-sync my enlistments, copy/edit over unsubmitted changes, and I should be good to go.


Right now, I am very glad that a long time ago I segregated all my source code enlistments onto one partition and separate from my OS or my own “My Document” partitions – recovery and backup procedures are very clear-cut depending on the partition – I don’t back up the OS and just reinstall whenever necessary; I backup and restore the “My Document” partition; I don’t backup the source code partition and re-sync whenever necessary.


So, I will be busy recovering from this ordeal when the replacement hard drive arrives and until then, I am going to be handicapped from doing productive development at work (I don’t consider email “productive” work). Maybe I will get time to spend on blogging, newsgroup, or investigating Sharepoint questions. Or maybe I will play around with native/managed Interop on IIS7 and post some code on that. 🙂


Oh, the choices, choices…


//David

Comments (6)

  1. Nicholas says:

    I often see a lot of tech-savvy people using Spinrite.. and I have to wonder.. why? It is a terrible program, just ask anyone who really repairs hard drives for a living. There’s just no way through software to magically fix a disk that is dead or on the verge of being dead.

  2. personally, i would invest in a SATA RAID solution for your desktop to avoid this in the future. 😉  

  3. David.Wang says:

    Nicholas – true, I do not expect software to fix a hardware malfunction. If a IC short-circuits or things like that, I do not think software helps.

    However, I do expect that higher data density and drive towards lower costs per platter must result in lower quality, and that is where proactive "maintenance" on the software side to notice "shady" hardware slip-ups should help.

    It is only a bonus if it can do something about the more catestrophic cases between "shady" behavior and "verge of being dead", but in those cases I’d go for the professional data recovery technicians.

    //David

  4. David.Wang says:

    Mathew – tell that to my manager! The outage in productivity is probably worth the SATA Raid solution. 🙂

    //David

  5. Nicholas says:

    I purchased my new Dell with the "Datasafe" option, came with two 160gig SATA drives in RAID1 via the on-board Intel SATA chipset. Works lovely, and Dell gave it to me for only like $50 more.

  6. David.Wang says:

    Nicholas – yeah, with the continuing falling price of hard drives, it is very much feasible to just do RAID for simple data protection. I’m getting a replacement 250G HDD for $90. The cost of me not being able to do any productive work the past four days – way more than $90.

    //David