[BBLISA] Whatever happened to Seagate?

[Nahum Shalman]

On Fri, May 8, 2015 at 7:47 AM, Edward Ned Harvey (bblisa4) <
bblisa4 at nedharvey.com> wrote:

> > From: bblisa [mailto:bblisa-bounces at bblisa.org] On Behalf Of Nahum
> > Shalman
> >
> > Just make sure you regularly apply power... Apparently SSDs can start
> losing
> > data in as little as three months without power:
> > https://twitter.com/bsdphk/status/596302795613011969  -
> > >   https://blog.korelogic.com/blog/2015/03/24#ssds-evidence-storage-
> > issues
>
> Interesting -
>
> It's interesting in that it's so poorly written and you accept it without
> question.


I do not accept it without question. However, until reading that article it
hadn't even occurred to me that putting an SSD on a shelf could result in
data degradation that quickly (I would consider 12 months to be quickly.)


> If there is some reality to the idea of needing to apply power to SSD's
> occasionally, I would love to hear it, but here are my comments for now:
>
> First off, SSD's use flash memory. Unlike DRAM, flash memory does not have
> a refresh cycle - so even if you power on the drive - nothing is going to
> change in the storage cell. My point is that powering on the drive won't
> actually *do* anything. We know the flash memory has a shelf life
> (everything does) but powering on the drive won't change that. If you're
> concerned, you might have to read the whole contents of the drive, and then
> write it all back again to the drive; THAT would actually refresh the
> memory cells. Not merely powering it on and leaving it idle and then
> powering it off.
>

Sure. My comment of "make sure you regularly apply power" illustrates
little more than my own ignorance of how the flash cells do or do not lose
state over time. Whatever the shelf life, my only concern was that anyone
truly ditching HDDs in favor of SSDs should be aware of the differences and
take the appropriate precautions accordingly. You make a good point that a
ZFS scrub would be far superior to simply applying power, though perhaps
even that wouldn't be sufficient if they need fresh writes (an even scarier
possibility.)


> The article referenced sounds techy and smart. And I'm not disputing that
> SSD's have shelf life (everything does.) And I'm not disputing that changes
> in temperature affect their shelf life (I haven't checked; but probably
> true.) But the sentence "They need consistent access to a power source in
> order for them to not lose data over time" and "A stored SSD, without
> power, can start to lose data" are literally the only sentences that say
> anything about power being required. They then go on to talk about
> temperature and environmental factors, but they never offer any explanation
> as to why power would be necessary, or how it would mitigate any problems.
>

It all merits research and/or experimentation. I just don't want to do it
with my data. The article came to my attention on the same day you wrote
about abandoning HDDs entirely in favor of SSDs and I thought anyone
following this thread would find it interesting.

-Nahum
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.bblisa.org/pipermail/bblisa/attachments/20150508/2b0ad2a2/attachment.html>