Discussion:
Is there a rotary hd size vs lifetime dependence?
(too old to reply)
Martha Adams
2020-01-09 17:31:13 UTC
Permalink
Do laptop size hd's wear out and fail after fewer service hours
than full-box size hd's? I can think about this and arrive at
conclusion that for today's state of the art making these things,
laptops hds and full-bix hds, have similar expectable lifetimes.
But -- what is observed reality? What is the experience of
observant users??

Titeotwawki -- Martha Adams [Thr 2020 Jan 09]
Jim Diamond
2020-01-09 19:59:11 UTC
Permalink
Post by Martha Adams
Do laptop size hd's wear out and fail after fewer service hours
than full-box size hd's? I can think about this and arrive at
conclusion that for today's state of the art making these things,
laptops hds and full-bix hds, have similar expectable lifetimes.
But -- what is observed reality? What is the experience of
observant users??
I have laptops over 10 years old that are still going strong, and
another one (Fujitsu T900) that will be 10 in October, and has
received lots of use (and the occasional accidental dropping). I've
replace the disks in most of these machines at least once, usually to
get more capacity, but occasionally because a disk rolled over and
died.

My oldest desktop in (continuous) use will be 10 in a few months.
Again, I have replaced some disks in that (and other desktops), but
usually for increased capacity, and only occasionally for failure (or
imminent failure predicted by smartctl output).

If you are considering mechanical disks, as opposed to solid state
disks (and equivalent), their enemy is mechanical shocks and being
moved while in operation. And, of course, disks in laptops are more
likely to be moved around while spinning than desktop disks. The only
disk that has failed me (before I replaced it) was in a laptop, but
you shouldn't read too much into one isolated incident.

For desktop disks, you might find it interesting to read
https://www.backblaze.com/blog/hard-drive-stats-q2-2019/. The web
page starts with "As of June 30, 2019, Backblaze had 110,640 spinning
hard drives in our ever-expanding cloud storage ecosystem."
These guys have lots of reliability data, but only (IIRC) for desktop
disks.

That doesn't answer your question. But I think both laptop and
desktop disks come in "this is a quality device which will last a long
time" and "this is not likely going to live for a long time"
varieties. The problem is, things keep changing.

I've had SSDs in laptops and desktops for a few years, and so far I've
had no problems. But that time frame is short.

Cheers.
John Forkosh
2020-01-10 10:11:32 UTC
Permalink
Post by Jim Diamond
Post by Martha Adams
Do laptop size hd's wear out and fail after fewer service hours
than full-box size hd's? I can think about this and arrive at
conclusion that for today's state of the art making these things,
laptops hds and full-bix hds, have similar expectable lifetimes.
But -- what is observed reality? What is the experience of
observant users??
I have laptops over 10 years old that are still going strong, and
another one (Fujitsu T900) that will be 10 in October, and has
received lots of use (and the occasional accidental dropping). I've
replace the disks in most of these machines at least once, usually to
get more capacity, but occasionally because a disk rolled over and
died.
My oldest desktop in (continuous) use will be 10 in a few months.
Again, I have replaced some disks in that (and other desktops), but
usually for increased capacity, and only occasionally for failure (or
imminent failure predicted by smartctl output).
If you are considering mechanical disks, as opposed to solid state
disks (and equivalent), their enemy is mechanical shocks and being
moved while in operation. And, of course, disks in laptops are more
likely to be moved around while spinning than desktop disks. The only
disk that has failed me (before I replaced it) was in a laptop, but
you shouldn't read too much into one isolated incident.
For desktop disks, you might find it interesting to read
https://www.backblaze.com/blog/hard-drive-stats-q2-2019/. The web
page starts with "As of June 30, 2019, Backblaze had 110,640 spinning
hard drives in our ever-expanding cloud storage ecosystem."
These guys have lots of reliability data, but only (IIRC) for desktop
disks.
That doesn't answer your question. But I think both laptop and
desktop disks come in "this is a quality device which will last
a long time" and "this is not likely going to live for a long time"
varieties. The problem is, things keep changing.
I've had SSDs in laptops and desktops for a few years,
and so far I've had no problems. But that time frame
is short. Cheers.
Also, regardless of mtbf's, etc, be aware of s.m.a.r.t.,
disk Self-Monitoring, Analysis and Reporting Technology...
https://en.wikipedia.org/wiki/S.M.A.R.T.

Just last summer I had a 3.5" 2TB five-year-old seagate
drive fail. But about a week before it completely died,
the post started reporting "smart" errors every time
I booted the box. I didn't know what that meant at the
time, so googled it for the bad news. But that gave me
enough lead time to do several exhaustive final sets
of backups. Meanwhile, reads kept taking longer and
longer as parity/whatever checks failed more and more
often. But no data at all was ultimately lost.

So, whether 2.5" or 3.5", I'd definitely recommend you
make sure that whatever disk you buy is s.m.a.r.t.
--
John Forkosh ( mailto: ***@f.com where j=john and f=forkosh )
John Forkosh
2020-01-10 10:27:13 UTC
Permalink
Post by Martha Adams
Do laptop size hd's wear out and fail after fewer service hours
than full-box size hd's? <<snip>>.....
..... regardless of mtbf's, etc, be aware of s.m.a.r.t.,
disk Self-Monitoring, Analysis and Reporting Technology...
https://en.wikipedia.org/wiki/S.M.A.R.T.
Just last summer I had a 3.5" 2TB five-year-old seagate
drive fail. But about a week before it completely died,
the post started reporting "smart" errors every time
I booted the box. I didn't know what that meant at the
time, so googled it for the bad news. But that gave me
enough lead time to do several exhaustive final sets
of backups. Meanwhile, reads kept taking longer and
longer as parity/whatever checks failed more and more
often. But no data at all was ultimately lost.
So, whether 2.5" or 3.5", I'd definitely recommend you
make sure that whatever disk you buy is s.m.a.r.t.
I forgot to mention the linux smartctl utility installed
with slackware. man smartctl for info, and just type
bash-5.0# smartctl -i /dev/sda (sudo or run as root)
for info about the corresponding disk. My 6TB replacement
for that failed 2TB gives...

bash-5.0# smartctl -i /dev/sdb
smartctl 7.0 2018-12-30 r4883 [x86_64-linux-4.19.30] (local build)
Copyright (C) 2002-18, Bruce Allen, Christian Franke, www.smartmontools.org

=== START OF INFORMATION SECTION ===
Model Family: Seagate Barracuda Pro
Device Model: ST6000DM004-2EH11C
Serial Number: ZA1EMBSP
LU WWN Device Id: 5 000c50 0b4e8a925
Firmware Version: DN04
User Capacity: 6,001,175,126,016 bytes [6.00 TB]
Sector Sizes: 512 bytes logical, 4096 bytes physical
Rotation Rate: 7200 rpm
Form Factor: 3.5 inches
Device is: In smartctl database [for details use: -P show]
ATA Version is: ACS-3 T13/2161-D revision 5
SATA Version is: SATA 3.1, 6.0 Gb/s (current: 6.0 Gb/s)
Local Time is: Fri Jan 10 05:19:23 2020 EST
SMART support is: Available - device has SMART capability.
SMART support is: Enabled
--
John Forkosh ( mailto: ***@f.com where j=john and f=forkosh )
Eric Pozharski
2020-01-10 11:35:06 UTC
Permalink
Post by Martha Adams
Do laptop size hd's wear out and fail after fewer service hours
than full-box size hd's? <<snip>>.....
*SKIP*
So, whether 2.5" or 3.5", I'd definitely recommend you make sure that
whatever disk you buy is s.m.a.r.t.
I forgot to mention the linux smartctl utility installed with
slackware. man smartctl for info, and just type
bash-5.0# smartctl -i /dev/sda (sudo or run as root)
Just don't panic when you do 'smartctl -A /dev/whatever'. When I've
seen this table for first time I'd literally shitted bricks.

*CUT*
--
Torvalds' goal for Linux is very simple: World Domination
Stallman's goal for GNU is even simpler: Freedom
Javier
2020-01-10 15:26:36 UTC
Permalink
Post by Martha Adams
Do laptop size hd's wear out and fail after fewer service hours
than full-box size hd's? I can think about this and arrive at
conclusion that for today's state of the art making these things,
laptops hds and full-bix hds, have similar expectable lifetimes.
But -- what is observed reality? What is the experience of
observant users??
Another thing to consider is the capacity of the harddisk. More Tb
means more plates and more heat and noise, and that possibly decreases
the lifetime of the HD, the battery, the psu and the computer. That
is something to consider in laptops and desktop computers with compact
form factors (ie tabletop pcs).

Bigger specs does not necessarily mean better. I would say that a 1
Tb disk and a 2 Tb disk from the same maker do not have the same
lifetimes.

You can buy an HD with a lower spinning speed (5400rpm instead of 7200rpm)
to decrease the power comsumption (at the cost of incresing access times).
Pascal Hambourg
2020-01-10 20:46:11 UTC
Permalink
Post by Javier
You can buy an HD with a lower spinning speed (5400rpm instead of 7200rpm)
to decrease the power comsumption (at the cost of incresing access times).
And decreased sequential read/write speed.
Martha Adams
2020-01-11 03:16:02 UTC
Permalink
Post by Pascal Hambourg
Post by Javier
You can buy an HD with a lower spinning speed (5400rpm instead of 7200rpm)
to decrease the power comsumption (at the cost of incresing access times).
And decreased sequential read/write speed.
====================================================================

Interesting responses here, especially re the smartctl resource. I'm
going to be looking in to that and may post here later re what I find.
But I notice that among responses here, none of those responds to my
question. Namely:

Is there a rotary hd size vs lifetime dependence?

Surely somewhere, there's something about that around in cyberspace,
and I'd think Slackers here would be the best kind of people to have
seen it.

This is a growable topic, since it can lead into discussion of hardware
hd's vs SSDs, which surely interests many besides myself. But right
now my interest is my rotary hd size topic.

Titewotwawki -- Martha Adams [Fri 2020 Jan 10]
Rich
2020-01-11 03:53:09 UTC
Permalink
Post by Martha Adams
Post by Pascal Hambourg
Post by Javier
You can buy an HD with a lower spinning speed (5400rpm instead of 7200rpm)
to decrease the power comsumption (at the cost of incresing access times).
And decreased sequential read/write speed.
====================================================================
Interesting responses here, especially re the smartctl resource. I'm
going to be looking in to that and may post here later re what I find.
But I notice that among responses here, none of those responds to my
Is there a rotary hd size vs lifetime dependence?
Besides anecdotal response, unless someone somewhere has done a
controlled study, you are unlikely to actually obtain a real answer.

And the one place that has done a real study (Backblaze) has studied
only 3.5 inch drive sizes, and so the results relate only to those
drives.
Post by Martha Adams
Surely somewhere, there's something about that around in cyberspace,
and I'd think Slackers here would be the best kind of people to have
seen it.
I've not seen it. Does not mean it does not exist, but there's one
vote.

Now, think logically about the issue. What is it that 'wears out' in a
mechanical hard drive? The moving parts, obviously. More
specifically, the bearing surfaces in the bearings that allow for the
moving parts to actually move.

Now think about what factors contribute to the useful lifetime of a
bearing:

1) quality of the bearing material
2) quality of manufacture of the bearing surfaces (i.e., smoothness
level)
3) amount of lubricant provided to the bearing
4) quality of that lubricant
5) useful lifetime of that lubricant
6) etc.

Now, if one builds a large drive bearing, and uses, say, a low quality
lubricant, and builds a corresponding small drive bearing with a very
high quality lubricant, with all other factors being equal, one would
expect the small drive bearing to outlast the large drive one.

Reverse the lubricants above, and one would expect the large drive
bearing to outlast the small drive one.

Which comes to my point... There are enough variables involved in
quality of materials and workmanship that one can't say "X size drive
would always outlast Y size drive" because 'outlast' is determined by
how well built each one was. And "size of drive" has no relationship
to "quality of materials and workmanship" of bearings inside that
drive.
Post by Martha Adams
This is a growable topic, since it can lead into discussion of hardware
hd's vs SSDs, which surely interests many besides myself.
SSD's have a wholly different 'wear out' aspect that is unrelated to
the causes of failures in mechanical drives.
John Forkosh
2020-01-11 09:07:31 UTC
Permalink
Post by Martha Adams
Post by Pascal Hambourg
Post by Javier
You can buy an HD with a lower spinning speed (5400rpm instead of 7200rpm)
to decrease the power comsumption (at the cost of incresing access times).
And decreased sequential read/write speed.
====================================================================
Interesting responses here, especially re the smartctl resource. I'm
going to be looking in to that and may post here later re what I find.
But I notice that among responses here, none of those responds to my
Is there a rotary hd size vs lifetime dependence?
Surely somewhere, there's something about that around in cyberspace,
and I'd think Slackers here would be the best kind of people to have
seen it.
This is a growable topic, since it can lead into discussion of hardware
hd's vs SSDs, which surely interests many besides myself. But right
now my interest is my rotary hd size topic.
Titewotwawki -- Martha Adams [Fri 2020 Jan 10]
I don't know the actual answer, but I'd wager it's not very useful
in the following sense:
The standard deviation of mtbf's among individual drives
is much greater than
the difference between the 2.5" mtbf and the 3.5" mtbf.

And ssd's vs hd's is a whole other issue. When I replaced my failed
2TB hd, I not only got a 6TB hd replacement, but also a 1TB ssd
for the box's /dev/sda boot. It's only an either/or choice for
laptops (and some support both an m2 and 2.5"), and in that case
ssd's aren't as mechanically sensitive as hd's, which should probably
be the deciding factor for most people.
--
John Forkosh ( mailto: ***@f.com where j=john and f=forkosh )
Loading...