Encryption effecting drive reliability

[Dennis the Nerf Herder]

Does anyone have experience with hard drive encryption reducing or
otherwise effecting the reliability of their hard drive?

While working with an enterprise worth of mixed desktop and laptop
computers recently, it seems the laptop have a higher rate of "issues"
since having the encryption installed. The desktop computers, many
with SATA drives, are not coming back to the service center nearly as
often as the laptops. At this point I am not in a position to
characterize whether I think it is the encryption software, one
particular type of drive, an operating system dependency of some type,
or (possibly) user error.

But I have downloaded several hard drive performance tools and run a
couple of drive manufacturer specific ones. I was very disappointed
with the results generated by the latter because all they gave was a
'result code' which had very little meaning. It did not make a
difference whether I ran the "Quick" test or "Advanced" test, the only
difference there was one took ten times longer to finish.

So I am seeking a utility that I can load under Microsoft Windows XP
and use to measure a hard drive's performance in BOTH its normal
(decrypted) and encrypted states. Particularly I would like to see if
there are increased "Read" or "Write" errors, or if the drive fails to
respond to a command (such as seeking to a specific sector - which may
only show as a re-attempt or something), then I will encrypt it and
check the performance again. And finally I will repeat the entire
process to determine if the results can be duplicated with some
consistency, or if maybe they are a rare occurance somehow tied to
either the drive or the encryption software.

I would love to recieve a reply from anyone that has done something
like this already, or who has knowledge of a detailed hard drive
performance monitoring program. I have tried "Hard Disk Sentinel"
(www.hdesentinel.hu) and "Drive Health" (by Helexis) and was
disappointed with their built-in features; they are fine just not what
I was hoping to see.

Thanks in advance!

 

[Rod Speed]

Does anyone have experience with hard drive encryption reducing
or otherwise effecting the reliability of their hard drive?

Nope, no one has. Some may have fooled themselves and decided that.

While working with an enterprise worth of mixed desktop and laptop
computers recently, it seems the laptop have a higher rate of "issues"
since having the encryption installed. The desktop computers, many
with SATA drives, are not coming back to the service center nearly as
often as the laptops.

That isnt surprising, laptops always have a harder life, get
belted around a lot more and the drives arent cooled as well.

At this point I am not in a position to characterize whether I think
it is the encryption software, one particular type of drive, an
operating system dependency of some type, or (possibly) user error.

Or just how laptops get used and their design limitations.

But I have downloaded several hard drive performance tools and run
a couple of drive manufacturer specific ones. I was very disappointed
with the results generated by the latter because all they gave was
a 'result code' which had very little meaning. It did not make a
difference whether I ran the "Quick" test or "Advanced" test, the
only difference there was one took ten times longer to finish.

You get more information from something like the Everest SMART report.

So I am seeking a utility that I can load under Microsoft Windows
XP and use to measure a hard drive's performance in BOTH its
normal (decrypted) and encrypted states.

Waste of time and wont tell you a damned thing even if you find one.

Particularly I would like to see if there
are increased "Read" or "Write" errors,

Any automated file copier that works at the XP level
combined with the drive SMART data will tell you that.

or if the drive fails to respond to a command (such as seeking
to a specific sector - which may only show as a re-attempt or
something), then I will encrypt it and check the performance
again. And finally I will repeat the entire process to determine
if the results can be duplicated with some consistency,

It wont be.

or if maybe they are a rare occurance somehow
tied to either the drive or the encryption software.

Thats what the SMART test does with the drive.

I would love to recieve a reply from anyone that has done
something like this already, or who has knowledge of a
detailed hard drive performance monitoring program.
I have tried "Hard Disk Sentinel" (www.hdesentinel.hu) and
"Drive Health" (by Helexis) and was disappointed with their
built-in features; they are fine just not what I was hoping to see.

What you are hoping to see is just plain silly with encryption.

 

[Arno Wagner]

Does anyone have experience with hard drive encryption reducing or
otherwise effecting the reliability of their hard drive?

While working with an enterprise worth of mixed desktop and laptop
computers recently, it seems the laptop have a higher rate of "issues"
since having the encryption installed. The desktop computers, many
with SATA drives, are not coming back to the service center nearly as
often as the laptops. At this point I am not in a position to
characterize whether I think it is the encryption software, one
particular type of drive, an operating system dependency of some type,
or (possibly) user error.

But I have downloaded several hard drive performance tools and run a
couple of drive manufacturer specific ones. I was very disappointed
with the results generated by the latter because all they gave was a
'result code' which had very little meaning. It did not make a
difference whether I ran the "Quick" test or "Advanced" test, the only
difference there was one took ten times longer to finish.

So I am seeking a utility that I can load under Microsoft Windows XP
and use to measure a hard drive's performance in BOTH its normal
(decrypted) and encrypted states. Particularly I would like to see if
there are increased "Read" or "Write" errors, or if the drive fails to
respond to a command (such as seeking to a specific sector - which may
only show as a re-attempt or something), then I will encrypt it and
check the performance again. And finally I will repeat the entire
process to determine if the results can be duplicated with some
consistency, or if maybe they are a rare occurance somehow tied to
either the drive or the encryption software.

I would love to recieve a reply from anyone that has done something
like this already, or who has knowledge of a detailed hard drive
performance monitoring program. I have tried "Hard Disk Sentinel"
(www.hdesentinel.hu) and "Drive Health" (by Helexis) and was
disappointed with their built-in features; they are fine just not what
I was hoping to see.

Thanks in advance!

For error monitoring, you basically need a tool that reads the
SMART status. It ir really difficult to get more from a drive
today. Personally I use the smartmontools (cmdline, also
available for windows).

What type of encryption are you talking about? I would be
surprised if the encryption itself was a problem. If, however,
it is badly designed or implemented, it could increase the
number of drive accesses and/or seeks, thereby increasing
the power consumption and drive temperature. That could very
well shorten drive lifetime.

A data-point: dm-crypt under Linux encrypts a whole device but
does not cause any change in disk access pattern, hence cannot
cause increased drive failure.

Arno

 

[Dennis the Nerf Herder]

Rod,

I have seen the SMART drive information and it did not
seem to be telling me very much about the drive either.
I will have to look at it again, but the numbers it
reported did not change even after several hours of
operation. So that leads me to believe either the drive
performance is very consistent or that the SMART data
reading utility MAY NOT be doing its job correctly.

It was my impression that SMART technology only
give an idea of the drive's lifespan and how far along
it is on the road to failure.

So far I have seen some measurable differences in
performance of the decrypted and encrypted hard
drive, however I am not ready to report the findings yet.
Someone else uploaded a report showing a few
different encryption programs and their relative
performance results. Some where 10% worse, 17%
worse and so on and so forth. I thought it was useful
even if the brands of encryption being reported did not
include the one I am using.

I always reserve judgement until ALL the facts of the
matter are considered.

 

[Dennis the Nerf Herder]

I am interested in "performance errors" not circuit malfunctions,
hours of use, temperature or whatever, because those are the
"physical" (electrical / electronic) statistics of the drive.

What I am thinking about is the performance of Windows TO
the encryption software TO the hard drive driver (?) TO the
BIOS then TO the hard drive itself. For example, if the drive
has write caching and the rate at which data is being pumped
into the cache buffer (thingy) is slowed down, then overall disk
performance is (probably) effected as well. Because I have
been told the data (programs, OS files, etc) is encrypted while
"at rest" so it MUST be going through the extra step of being
decrypted before it can be used.

Some more experienced technicians may counter with, 'Well
the "driver" or whatever the encryption software is so fast its
effect is negligible. But if that is the case why am I seeing
many more laptops with "typical" use having problems?
It seems like the hardware is indicated as possibly having
difficulty interacting with the WHOLE process.

Again, I have to write that I am reserving judgement until I can
find out more information.

The software is a full disk encryption program made by Guard-
ian Edge but I would like to withhold the version number
currently being used.

Arno, thanks for your reply.

 

[Rod Speed]

I have seen the SMART drive information and it did not
seem to be telling me very much about the drive either.

Its only meant to report problems.

I will have to look at it again, but the numbers it reported
did not change even after several hours of operation.

Not surprising if the drive is fine.

So that leads me to believe either the drive performance
is very consistent or that the SMART data reading utility
MAY NOT be doing its job correctly.

Or you dont actually understand what SMART is about.

It was my impression that SMART technology only give an idea
of the drive's lifespan and how far along it is on the road to failure.

Its intended to record problems seen by the drive and give some
indication if that is likely an indication of imminent drive failure.

Its never intended to say anything about the life
of a drive which hasnt exhibited any problems.

So far I have seen some measurable differences in
performance of the decrypted and encrypted hard
drive, however I am not ready to report the findings yet.

Thats nothing like your original about RELIABILITY.

Someone else uploaded a report showing a few different
encryption programs and their relative performance results.
Some where 10% worse, 17% worse and so on and so forth.

Thats nothing like your original about RELIABILITY.

I thought it was useful even if the brands of encryption
being reported did not include the one I am using.

I always reserve judgement until ALL
the facts of the matter are considered.

Easy to claim. You dont appear to even be able to work out
the difference between RELIABILITY and PERFORMANCE.

 

[Arno Wagner]

I am interested in "performance errors" not circuit malfunctions,
hours of use, temperature or whatever, because those are the
"physical" (electrical / electronic) statistics of the drive.

What I am thinking about is the performance of Windows TO
the encryption software TO the hard drive driver (?) TO the
BIOS then TO the hard drive itself. For example, if the drive
has write caching and the rate at which data is being pumped
into the cache buffer (thingy) is slowed down, then overall disk
performance is (probably) effected as well. Because I have
been told the data (programs, OS files, etc) is encrypted while
"at rest" so it MUST be going through the extra step of being
decrypted before it can be used.

No. All data is stopped (gathered into larger portions) in the OS
write buffer anyways. It is then written at the eraliest convenience
or when aged enough. Don't know what timeout Windows uses, but
Linux and Novell (historic) use 300 seconds for the hard flush per
default.

Some more experienced technicians may counter with, 'Well
the "driver" or whatever the encryption software is so fast its
effect is negligible.

Not quite ture. the wauai is so long in the normal case, that
the encryption is fast enough to be slower... ;-)

But if that is the case why am I seeing
many more laptops with "typical" use having problems?
It seems like the hardware is indicated as possibly having
difficulty interacting with the WHOLE process.

First this may be an unreleated effect. Secondly, maybe becasue
this encryption is badly implemented. It may re-write whole file on
changes ot the like and so indeed cause more accesses. That
was why I asked what kind of encryption you are talking about...

Again, I have to write that I am reserving judgement until I can
find out more information.

The software is a full disk encryption program made by Guard-
ian Edge but I would like to withhold the version number
currently being used.

Hmm. One thing you can do is comparte the drive temperatures
of the same drive type being run with and without encryption.
Maybe log it for a few days of normal use. If the encrypted
drives see more use, then their temperatures should be higher.
Also higher temperature is the only ageing-acceletarion contition
I know for HDDs. The number of writes to not really matter...

Arno, thanks for your reply.

No problem. If you find out something, please post your results
here, I would like to knoe.

Arno

 

[Rod Speed]

I am interested in "performance errors"

Whatever the hell those are supposed to be.

not circuit malfunctions, hours of use, temperature
or whatever, because those are the "physical"
(electrical / electronic) statistics of the drive.

What I am thinking about is the performance of Windows
TO the encryption software TO the hard drive driver (?)
TO the BIOS then TO the hard drive itself. For example,
if the drive has write caching and the rate at which data is
being pumped into the cache buffer (thingy) is slowed down,
then overall disk performance is (probably) effected as well.
Nope.

Because I have been told the data (programs,
OS files, etc) is encrypted while "at rest"

Whatever the hell that is supposed to be.

so it MUST be going through the extra step
of being decrypted before it can be used.
Duh.

Some more experienced technicians may counter with,
'Well the "driver" or whatever the encryption software is
so fast its effect is negligible.

And they'd be right.

But if that is the case why am I seeing many
more laptops with "typical" use having problems?

Just because laptops are much more roughly treated
and dont get anywhere near as well cooled as desktops.

It seems like the hardware is indicated as possibly
having difficulty interacting with the WHOLE process.

Then you need to get your seems machinery seen to.

Again, I have to write that I am reserving
judgement until I can find out more information.

You have always been, and always will be, completely
and utterly irrelevant. Your 'judgement' in spades.

The software is a full disk encryption program
made by Guardian Edge but I would like to
withhold the version number currently being used.

Pathetic, really.

 

[Dennis the Nerf Herder]

Whatever the hell those are supposed to be.
{snip}

Whatever the hell that is supposed to be.


And they'd be right.


Just because laptops are much more roughly treated
and dont get anywhere near as well cooled as desktops.


Then you need to get your seems machinery seen to.


You have always been, and always will be, completely
and utterly irrelevant. Your 'judgement' in spades.


Pathetic, really.

Your replies have been overly critical, unhelpful and
irrelevant. So who is actually pathetic?

I appologize to the Group for using this small amount of
bandwidth to Reply to "Rod Speed" but was, in fact,
saddened and irritated by his eplies.

Didn't your parents teach you, 'If you can't say something
nice, don't say anything at all'? What are you, 14 years
old or something?

Over my years in technical support and information
technology, I have either received compliments and
thanks for my observations, support, feedback and input,
or NO COMMENTS AT ALL from clients and customers.

Rod, I invite you to keep your comments to yourself in
the future. Have a great day!

 

[Rod Speed]

Your replies have been overly critical, unhelpful and irrelevant.

Never ever could bullshit its way out of a wet paper bag.

So who is actually pathetic?

You, as always.

I appologize to the Group for using this small amount of bandwidth to
Reply to "Rod Speed" but was, in fact, saddened and irritated by his eplies.

You were in fact irritated by having your terminal stupiditys exposed.

Didn't your parents teach you, 'If you can't
say something nice, don't say anything at all'?

That aint what technical newsgroups are about, fool.

What are you, 14 years old or something?

Likely old enough to be your father, thanks.
Maybe even old enough to be your grandfather too.

Over my years in technical support and information
technology, I have either received compliments and
thanks for my observations, support, feedback and input,
or NO COMMENTS AT ALL from clients and customers.

Pity about your terminal stupiditys that have
been exposed for what they are in newsgroups.

Rod, I invite you to keep your comments to yourself in the future.

Gutless, I order you to go and **** yourself.

Have a great day!

Pathetic, really.