[kwlug-disc] OT: SSD disks?

[B.S.]

----- Original Message -----

> From: Rashkae <rashkae at tigershaunt.com>
> To: KWLUG discussion <kwlug-disc at kwlug.org>
> Cc: 
> Sent: Wednesday, November 11, 2015 11:59 AM
> Subject: Re: [kwlug-disc] OT: SSD disks?
> 
> On 15-11-10 02:23 PM, Chris Irwin wrote:
>>  On Tue, Nov 10, 2015 at 12:08 PM, B.S. <bs27975 at yahoo.ca> wrote:
>> 
>>>  Bear in mind the SSD warnings though - swapping increases SSD wear,
>> 
>>  I think worrying about swapping to an SSD isn't necessary now.
>> 
>>  Tech Report did an SSD endurance experiment with a few popular SSDs until
>>  they all failed. Spoiler: Don't buy
>> 
>> 
>> 
> http://techreport.com/review/27909/the-ssd-endurance-experiment-theyre-all-dead
>> 
>> 
> 
> Interesting tests academically, but I find myself supremely unconcerned 
> with how long past a drive's expected life you can torture it beyond any 
> reasonable use pattern before it implodes.
> 
> A more important test would be to stress test all the drives to the last 
> 10% of their rated life (as reported by SMART SSD Life, or equivalent), 
> then leave unpowered for 1 month.  The real test will be, which drives 
> will retain all their data after 1 month of no power with worn NAND.


I suspect not.

The real world scenario is - can one, and which, trust these newfangled SSD drives? And, aside from the is the price different worth it, especially vis a vis the purported increase in speed (anecdotal evidence here says no), is the price and capacity differential over spinning drives worth it / are they more/less reliable than said spinning drives?

(Putting aside the eco/power consumption elements, here. And never mind that avoiding single point of failure means you're going to have multiple/redundant storage, of whatever technology used.)

The torture tests speak to and reassure reliability, and ability to extract at end of life. Especially for the vast majority of systems where there is but this one continuously used, non-redundant disk. Most every system, especially laptops, have a single OS disk, including most desktops and non-cloud servers. (Not all, but I expect most.)

They note that warnings of imminent problems are well exposed, giving one time to address, to at least as good a level as spinning disks. Spinning or not, ignore such warnings at your peril.

Most systems could not / would not be left unpowered for a month, so what happens after a month, end of life or no, is largely irrelevant. To me the torture tests, speaking to the reliability and ability to recover, vis a vis spinning disks, is more important.


Now, if you want to discuss long term archival abilities, yes - but even then end of life issues largely become irrelevant. Archiving historical data on questionable media seems imprudent, regardless of the media.

People were backing up to optical media for said permanent storage that turned out not to be. (Optical media having dyes, and dyes fade ...) Along the way, the amount of data to be archived seems to have grown exponentially - thus some of the reason for the growth to the TB drives we see available today. Given the low SSD capacities vis a vis these TB drives, they do not seem to fit this purpose. Thus whether or not they retain their data, when we know the more cost effective spinning drives do, isn't so important to me.

Given the large amounts of data now to be backed up and archived, the only reasonable solution appears to be having redundant storage systems elsewhere on one's network. Cascading archives off site has become more intransigent given these capacities, and the failure of optical to keep up - both in capacity and speed. Double-density / double-sided blu-ray media is stupid expensive, SSDs may be worse, so SSD's aren't likely to be considered for the scenario.

Thus, I guess, cloud growth as off-site backup/archival storage. Which has its own excessive costs, and risks.