"The SSD Endurance Experiment: They're all dead"

"The SSD Endurance Experiment: They're all dead"
http://techreport.com/review/27909/t...heyre-all-dead

Why did Intel brick the 335?

Lynn

On 3/13/2015 5:28 PM, Lynn McGuire wrote:
"The SSD Endurance Experiment: They're all dead"
http://techreport.com/review/27909/t...heyre-all-dead

Why did Intel brick the 335?

Lynn

Why did the Intel 335 brick itself? That says to me that they do not have confidence in their checksums. Or, they are afraid of a
cascade failure (I’ve had several drives do that to me over the years).

Lynn

On 13/03/2015 8:35 PM, Lynn McGuire wrote:
On 3/13/2015 5:28 PM, Lynn McGuire wrote:
"The SSD Endurance Experiment: They're all dead"

http://techreport.com/review/27909/t...heyre-all-dead


Why did Intel brick the 335?

Lynn

Why did the Intel 335 brick itself? That says to me that they do not
have confidence in their checksums. Or, they are afraid of a cascade
failure (I’ve had several drives do that to me over the years).

Interesting experiment, one of my drives is approaching 14TB of writes
so far, still showing 99% of life left.

It seems that the Intel is the one that's the most conservative about
its lifespan calculations. Perhaps it's reasonable for a corporation
that prides itself on its reliability?

Yousuf Khan

On 14/3/15 1:31 PM, Yousuf Khan wrote:
It seems that the Intel is the one that's the most conservative about
its lifespan calculations. Perhaps it's reasonable for a corporation
that prides itself on its reliability?

Definitely better than being surprised and lost your data!

--
@~@ Remain silent. Nothing from soldiers and magicians is real!
/ v \ Simplicity is Beauty! May the Force and farces be with you!
/( _ )\ (Fedora release 21) Linux 3.18.8-201.fc21.i686+PAE
^ ^ 20:33:02 up 3 days 1:35 0 users load average: 0.01 0.04 0.05
�借貸! ��騙! ��交! �打交! �打劫! �自殺! 請考慮綜� (CSSA):
http://www.swd.gov.hk/tc/index/site_...sub_addressesa

On 3/14/2015 7:41 AM, Mr. Man-wai Chang wrote:
On 14/3/15 1:31 PM, Yousuf Khan wrote:
It seems that the Intel is the one that's the most conservative about
its lifespan calculations. Perhaps it's reasonable for a corporation
that prides itself on its reliability?

Definitely better than being surprised and lost your data!

Did you read the article? The Intel SSD cut off write access when it
reached its life and then bricked itself on the next reboot. I would
count that as a serious potential data loss.

Lynn

On Sat, 14 Mar 2015 17:31:53 -0500, Lynn McGuire
wrote:

On 3/14/2015 7:41 AM, Mr. Man-wai Chang wrote:
On 14/3/15 1:31 PM, Yousuf Khan wrote:
It seems that the Intel is the one that's the most conservative about
its lifespan calculations. Perhaps it's reasonable for a corporation
that prides itself on its reliability?

Definitely better than being surprised and lost your data!

Did you read the article? The Intel SSD cut off write access when it
reached its life and then bricked itself on the next reboot. I would
count that as a serious potential data loss.
Correct. What should be done instead is something like:
When 10% life is left enter a mode that requires read only access
from the outside. At that point the user would be allowed change
the free limit to 9% before requiring read only access, 8%, etc.
down to, say 5%. At that point only user reads are permitted.

There should be a small amount (say 1/20th) of write life retained to
handle while retiring the device by backing up the data. Some more
life is needed to handle "aging" data in case the device can't be
backed up.

Extra user action should be required to do data reads so
that an essentially read only device doesn't die by data decay.

Note that SMART data may be lost if the spare blocks really run
out.

Lynn

Lynn McGuire wrote:
On 3/14/2015 7:41 AM, Mr. Man-wai Chang wrote:
On 14/3/15 1:31 PM, Yousuf Khan wrote:
It seems that the Intel is the one that's the most conservative about
its lifespan calculations. Perhaps it's reasonable for a corporation
that prides itself on its reliability?

Definitely better than being surprised and lost your data!

Did you read the article? The Intel SSD cut off write access when it
reached its life and then bricked itself on the next reboot. I would
count that as a serious potential data loss.

Lynn

I agree. A sensible device will turn read-only at a time where
it can still ensure read-integrity for a reasonable time, but
it will _not_ brick itself. My guess is that SSDs these days are
still "experimental" and that we will need a decade or so before
sensible behavior is the norm.

Arno

Arno wrote:
Lynn McGuire wrote:
On 3/14/2015 7:41 AM, Mr. Man-wai Chang wrote:
On 14/3/15 1:31 PM, Yousuf Khan wrote:
It seems that the Intel is the one that's the most conservative about
its lifespan calculations. Perhaps it's reasonable for a corporation
that prides itself on its reliability?

Definitely better than being surprised and lost your data!

Did you read the article? The Intel SSD cut off write access when it
reached its life and then bricked itself on the next reboot. I would
count that as a serious potential data loss.

Lynn

I agree. A sensible device will turn read-only at a time where
it can still ensure read-integrity for a reasonable time, but
it will _not_ brick itself. My guess is that SSDs these days are
still "experimental" and that we will need a decade or so before
sensible behavior is the norm.

Arno

The Intel SSD on my Acer Aspire 1 did that; failed on a write, then
completely failed. Fortunately I take weekly backups and immediate
backups to a USB device of important stuff.

On 1/4/15 4:13 AM, Jerry Peters wrote:
The Intel SSD on my Acer Aspire 1 did that; failed on a write, then
completely failed. Fortunately I take weekly backups and immediate
backups to a USB device of important stuff.


You should never put data in a SSD, only re-installable programs and OS!

--
@~@ Remain silent. Nothing from soldiers and magicians is real!
/ v \ Simplicity is Beauty! May the Force and farces be with you!
/( _ )\ (Fedora release 21) Linux 3.19.3-200.fc21.i686+PAE
^ ^ 22:57:01 up 14:02 0 users load average: 0.09 0.12 0.10
�借貸! ��騙! ��交! �打交! �打劫! �自殺! 請考慮綜� (CSSA):
http://www.swd.gov.hk/tc/index/site_...sub_addressesa

Mr. Man-wai Chang wrote:
On 1/4/15 4:13 AM, Jerry Peters wrote:
The Intel SSD on my Acer Aspire 1 did that; failed on a write, then
completely failed. Fortunately I take weekly backups and immediate
backups to a USB device of important stuff.


You should never put data in a SSD, only re-installable programs and OS!

/ & /usr are on the SSD, /var & /home are on the SDHC card in the left
slot, & /tmp is a tmpfs. The SSD did last over 5 years BTW.

This is an Acer Aspire 1 *netbook*, the only choices are SSD & a SDHC
card.