3 ms·

I don't doubt your technical prowess for repairing HDDs, but my skepticism is based largely in that despite how overwhelmingly well known it is, it's never been

by csydas 10y ago

I don't doubt your technical prowess for repairing HDDs, but my skepticism is based largely in that despite how overwhelmingly well known it is, it's never been really tested in a formal setting and the explanations have never gone beyond just "expanding and contracting", with no deeper explanation. Aside from everyone's stories about how it worked, there really isn't a lot of good solid knowledge about whether or not it actually does anything. I'm not doubting stories where it went into the freezer and worked; quite the contrary, I'm more than willing to believe that was part of what was done to do it. What I question is whether or not it was the actual freezing process that did anything, if there actually is a physical change that occurs beyond people's speculation. It would seem to me that if there is a physical change going on with the drive, it should be visible, or at least a difference ought be visible in a before an after, save a few minor parts not visible to the naked eye or without further disassembly. The fact that there really hasn't been any research on this, even amateur, despite the number of server farms, kind of makes me wonder if it's not just a bit of old voodoo.

dogma1138 10y ago
Freezing (as well as sharp acceleration) can unstick the mechanical parts of the HDD quite often the arm that moves the heads up and down the platter. In my case i could hear the "ticks" of the arm stepper motor they were gone after a day in the freezer. And I've seen an iphone that would not work you put it into a freezer for 1-2 hours, take it it turns on, after about 1-2minutes it dies again and you could repeat it. There are probably so many things that temperature could affect when it comes to electronics that I'm not sure if there is even one thing that a research could easily focus on.
- csydas 10y ago
  Well, that's sort of my point - with how many different items "could" be studied and quantified, not a single study has been produced to measure even one aspect just to say "we noticed no discernible change after the freezing process". There isn't even really a good baseline statistic for what the success rate is. In the 30 some years I've been hearing the freezing idea, I find it really peculiar that nothing besides anecdotal evidence, not even amateur studies of just like a server farm testing failing drives or an IT desk recording their attempts that gives a sample size of more than ~10ish. Conjecture and theories are great, and it's good to hear that these drives came back online. But, without exaggeration, I've tried this on hundreds of drives over the course of my career when doing IT support, and I can think of maybe 2 instances when after the freeze, the drive spun up long enough for data t be recovered. And everyone's stories seem to be this way - just this one or two times, after a freeze, it worked. There are no stories where a company comes out and says "man, freezing drives has saved our butts and customer data tons of times!" I can't stress enough that this isn't a declaration that freezing doesn't do anything. But I feel the skepticism is absolutely appropriate given the voodoo qualities of the process and the overwhelming failure rate the process has. The lack of research on it and the fact that it's just a "hail mary" for most people rather than a cemented procedure just seems to me like the procedure is just ceremony, and that if put to any sort of actual trial, the success rate would fall well within the range of chance.

by csydas 10y ago

dogma1138 10y ago

Freezing (as well as sharp acceleration) can unstick the mechanical parts of the HDD quite often the arm that moves the heads up and down the platter. In my case i could hear the "ticks" of the arm stepper motor they were gone after a day in the freezer. And I've seen an iphone that would not work you put it into a freezer for 1-2 hours, take it it turns on, after about 1-2minutes it dies again and you could repeat it. There are probably so many things that temperature could affect when it comes to electronics that I'm not sure if there is even one thing that a research could easily focus on.

csydas 10y ago
Well, that's sort of my point - with how many different items "could" be studied and quantified, not a single study has been produced to measure even one aspect just to say "we noticed no discernible change after the freezing process". There isn't even really a good baseline statistic for what the success rate is. In the 30 some years I've been hearing the freezing idea, I find it really peculiar that nothing besides anecdotal evidence, not even amateur studies of just like a server farm testing failing drives or an IT desk recording their attempts that gives a sample size of more than ~10ish. Conjecture and theories are great, and it's good to hear that these drives came back online. But, without exaggeration, I've tried this on hundreds of drives over the course of my career when doing IT support, and I can think of maybe 2 instances when after the freeze, the drive spun up long enough for data t be recovered. And everyone's stories seem to be this way - just this one or two times, after a freeze, it worked. There are no stories where a company comes out and says "man, freezing drives has saved our butts and customer data tons of times!" I can't stress enough that this isn't a declaration that freezing doesn't do anything. But I feel the skepticism is absolutely appropriate given the voodoo qualities of the process and the overwhelming failure rate the process has. The lack of research on it and the fact that it's just a "hail mary" for most people rather than a cemented procedure just seems to me like the procedure is just ceremony, and that if put to any sort of actual trial, the success rate would fall well within the range of chance.