Trans OS X-Post: What do people do about obtaining non-shingled laptop drives?

[Java Jive]

Please excuse the Linux/Windows crosspost, this is a question about HD
hardware relevant to both OSs!

I have a Dell Precision M6300 that is slowing down really badly, and I
suspect, but have yet to prove, that the HD is failing.

It seems to be getting increasingly difficult to obtain non-shingled
replacement laptop drives. Samsung have sold out to Seagate, and
seemingly now most or all Seagate and Western Digital laptop drives are
SMR ...

Apparently the only non-shingled laptop drives currently made by Seagate
are Exos E, and v. expensive:

https://www.seagate.com/gb/en/products/cmr-smr-list/
https://www.ebuyer.com/store/Storage/cat/Hard-Drive---Internal?a00489=2.5%22&q=exos

Up-to-date information on WD drives seems irresponsibly hard to come by.
After the public backlash around 2020, lists were published then of
which WD drives were SMR ...

https://blog.westerndigital.com/wd-red-nas-drives/

... but that was 3 years ago and I've not found anything more up to date
and official from the WD site. Also, most independent lists are quite
old, dating from the time the scandal first broke, and/or are compiled
by NAS sites for desktop drives.

Of course, one could buy an older model drive very cheaply, but, even
when they have good ratings, at least some of the stock, even when new
- as in genuinely unused - have been on the shelf for so long that
they are already beyond manufacturer warranty, but, far too frequently,
are suspected items previously returned as faulty being resold, or just
plain second-hand/used and 'refurbished', whatever that may mean for an
item that has 'no user serviceable parts inside':

https://www.amazon.co.uk/Western-Digital-AV-GP-Intellipower-Internal-disk-disc-storage-gigabyte/product-reviews/B002P3KO7O/ref=cm_cr_arp_d_viewpnt_rgt?ie=UTF8&reviewerType=all_reviews&filterByStar=critical&pageNumber=1

Of course, that is a deliberately biased sample by looking at the
critical reviews, but I find them a useful measure of "What's the worst
that can happen?!"

Can anyone point to a UK source of reliable, genuinely new, moderately
priced non-shingled laptop drives from about 500GB to 1.5TB?

Of course, I could skip the shingles problem by going for an SSD, but
have not really explored this up til now. Experience and thoughts on
that would be welcome too.

TIA.

--

Fake news kills!

I may be contacted via the contact address given on my website:
www.macfh.co.uk

[Theo]

In uk.comp.os.linux Java Jive <ja...@evij.com.invalid> wrote:
> Can anyone point to a UK source of reliable, genuinely new, moderately
> priced non-shingled laptop drives from about 500GB to 1.5TB?
>
> Of course, I could skip the shingles problem by going for an SSD, but
> have not really explored this up til now. Experience and thoughts on
> that would be welcome too.

TBH there's little point in 2.5" HDD in that size range these days. The
cheapest and nastiest SSDs will perform better than any HDD. Frex:

512GB £23
https://www.ebuyer.com/1535248-patriot-p210-512gb-2-5-sata-iii-ssd-p210s512g25

1TB £40
https://www.ebuyer.com/1535247-patriot-p210-1tb-2-5-sata-iii-ssd-p210s1tb25

2TB £84
https://www.ebuyer.com/1535246-patriot-p210-2tb-2-5-sata-iii-ssd-p210s2tb25

Now I'd not be queuing up to buy those specific drives (I'd research and
likely spend a little more to get something better) but even these will be
night and day better than HDD.

Theo

[Andy Burns]

Java Jive wrote:

> I could skip the shingles problem by going for an SSD, but have not
> really explored this up til now.

I wouldn't fit anything other than an SSD to a laptop, seriously.

[J.O. Aho]

On 4/6/23 17:02, Java Jive wrote:
> Please excuse the Linux/Windows crosspost, this is a question about HD
> hardware relevant to both OSs!
>
> I have a Dell Precision M6300 that is slowing down really badly, and I
> suspect, but have yet to prove, that the HD is failing.
>

> Can anyone point to a UK source of reliable, genuinely new, moderately
> priced non-shingled laptop drives from about 500GB to 1.5TB?

I would recommend a SSD, no point in going for a HDD unless you need it
for large scale storage 2TB+

Here is my suggestion:
https://www.amazon.co.uk/Samsung-MZ-76E1T0B-EU-Solid-State/dp/B078WST5RK/

If you want to save some bucks, then go with the 500G, but try to avoid
250G SSD, they tend to be slower. I don't recommend QVO as it wears out
faster.

--
//Aho

[Char Jackson]

On Thu, 6 Apr 2023 16:57:42 +0100, Andy Burns <use...@andyburns.uk>
wrote:

+1

[Jim Kelly]

On 06/04/2023 17:11, J.O. Aho wrote:
>
> I don't recommend QVO as it wears out faster.

How fast? 5 years, 4 years, 3 years or just 6 months?

For most people if a hard disk lasts for 5 years then they have done
very well indeed. I have a HDD that has lasted for nearly 10 years but I
am not a 24/7 user of the machine. I switch on the machine once in the
evening, check the email in my private account, browse the web to see
what is in the news and that's all about it. It is then time to go to
bed after dinner to wake up in the morning to go to work.

Do you have a link where it says QVO is no better than EVO or is it just
your anecdotal experience of using different versions over the years.

Thank you for the info anyway.

[Sjouke Burry]

My HD is from 2004, or about 19 years.
80 GB , 25 GB used by XP PRO.
Never had any trouble.
All err info on HD reports OK.
Most data is on drive D.

[J.O. Aho]

On 4/6/23 19:53, Jim Kelly wrote:
> On 06/04/2023 17:11, J.O. Aho wrote:
>>
>> I don't recommend QVO as it wears out faster.
>
> How fast? 5 years, 4 years, 3 years or just 6 months?

Time depends on how much you write to the disk, this includes the
resizing of the paging file that windows does quite frequently in the
background. You can look at the product warranty for the 1T evo 600 TBW
(Max 5 years) vs 1T qvo 360 TBW (Max 3 years), that already hints that
Samsung expects the qvo to have shorter lifespan than the evo, sure this
number don't mean that all qvo will just last 3 years + 1 day or 360TB + 1B.

> For most people if a hard disk lasts for 5 years then they have done
> very well indeed. I have a HDD that has lasted for nearly 10 years but I
> am not a 24/7 user of the machine. I switch on the machine once in the
> evening, check the email in my private account, browse the web to see
> what is in the news and that's all about it. It is then time to go to
> bed after dinner to wake up in the morning to go to work.

I'm of the 24/7 school, I don't switch harddrives/ssd that often, but
when I do it's more to get more space and I do rather spend a few extra
bucks to get a better HDD/SSD and I tend to look at Backblaze yearly
report to feel safe with my pick.

> Do you have a link where it says QVO is no better than EVO or is it just
> your anecdotal experience of using different versions over the years.

https://www.digitaltrends.com/computing/samsung-qvo-vs-evo-guide/


--
//Aho

[Carlos E.R.]

On 2023-04-06 17:02, Java Jive wrote:
> Please excuse the Linux/Windows crosspost, this is a question about HD
> hardware relevant to both OSs!
>
> I have a Dell Precision M6300 that is slowing down really badly, and I
> suspect, but have yet to prove, that the HD is failing.

At least in Linux, this is easy to check. Assuming the drive is
/dev/sda, do, as root:

smartctl -a /dev/sda

Check these two lines:

ID# ATTRIBUTE_NAME FLAG VALUE WORST THRESH TYPE
UPDATED WHEN_FAILED RAW_VALUE


197 Current_Pending_Sector 0x0012 100 100 000 Old_age Always
- 0
198 Offline_Uncorrectable 0x0010 100 100 000 Old_age
Offline - 0


If the last column is not zero, you have a problem. Then look at this
other line:

5 Reallocated_Sector_Ct 0x0033 100 100 010 Pre-fail
Always - 0

If the two "100" are not 100, that's bad, the disk is dying. Replace it.

Go for an SSD, don't hesitate. Till 1TB at least the prices are reasonable.

I saw some laptops that had both SSD and rotating rust.


SSDs are actually more robust than traditional disks, they don't mind
vibrations, and are nicer on the battery.


--
Cheers, Carlos.

[David W. Hodgins]

On Thu, 06 Apr 2023 17:22:03 -0400, Carlos E.R. <robin_...@es.invalid> wrote:
> SSDs are actually more robust than traditional disks, they don't mind
> vibrations, and are nicer on the battery.

I did have a problem with one laptop when I tried replacing it's hard drive
with an ssd. Massive overheating during large writes (linux install) forcing
a system shutdown part way through. Put the old hard drive back in and it was
fine.

I ended up adding that ssd drive in my desktop system with a fan.
hddtemp for it shows ...
/dev/sdd: INTEL SSDSC2BW240A4: no sensor

It does get very hot to the touch even with the fan.

Other laptops I've put other ssd drives have been fine.

Regards, Dave Hodgins

[Carlos E.R.]

Yes, I suppose there were bad designs while the technology matured.

I replaced with an ssd the hard disk in my first laptop, a bit clunky
item for nowdays, and the thing boots when it wants. Sometimes it boots,
sometimes the computer thinks there is no disk. ctrl-alt-del and try
again. But absolutely no problems once booted.

I thought that maybe there is a faulty contact, maybe I should open the
laptop and reseat the cables. I have been postponing that for years and
years. Some day. Or maybe not. :-)

--
Cheers, Carlos.

[Martin Gregorie]

On Thu, 6 Apr 2023 23:22:03 +0200, Carlos E.R. wrote:

> On 2023-04-06 17:02, Java Jive wrote:
>> Please excuse the Linux/Windows crosspost, this is a question about HD
>> hardware relevant to both OSs!
>>
>> I have a Dell Precision M6300 that is slowing down really badly, and I
>> suspect, but have yet to prove, that the HD is failing.
>
> At least in Linux, this is easy to check. Assuming the drive is
> /dev/sda, do, as root:
>
> smartctl -a /dev/sda
>
> Check these two lines:
>
> ID# ATTRIBUTE_NAME FLAG VALUE WORST THRESH TYPE UPDATED
> WHEN_FAILED RAW_VALUE
>

I like to keep an eye on disk ages and potential problems, so I have a
weekly cronjob that runs smartctl and emails me the report it produces.

I've found that spinning rust tends to fail after an accumulated running
time of around 50,000 hours +/-5,000

I usually fit 500GB WD Blue drives, which don't use shingling or other
similar tech. I've had no problems with either 2.5" or 3.5" drives for a
long time now.

I'm also happy with the Sanyo 120GB SSD I've fitted in a Lenovo R61i when
its original Fujitsu hard disk died.

This machine's disk access hardware can't handle disks bigger than 200GB,
and by the time its original 160 GB HDD died you couldn't buy any HDDs
smaller than 320GB, so fitting a 120GB SSD was the obvious answer. It
certainly goes like the clappers with this installed. It also produces
weekly smartctl reports: the most notable difference, apart from a general
speedup compared with spinning rust, is that the active hours per week
figure is a lot lower than for spinning rust, probably because an SSD is
instant on/instant off while an HDD will include spinup, spindown and
idle_but_spunup time in its accumulated active runtime.

So far, that's my only experience with an SSD, but my ancient Dual Athlon
house server recently died horribly and is being replaced by a new box
containing a 1TB WD SSD, so I'll be interested to see how this storage
works out. I'll keep the weekly smartctl reports comming on its
accumulated running time. .



--

Martin | martin at
Gregorie | gregorie dot org

[Java Jive]

On 06/04/2023 16:02, Java Jive wrote:
>
> Please excuse the Linux/Windows crosspost, this is a question about HD
> hardware relevant to both OSs!

Thanks for all the replies, all of which I've read and noted.

> I have a Dell Precision M6300 that is slowing down really badly, and I
> suspect, but have yet to prove, that the HD is failing.

This afternoon I got around to testing the HD with CrystalMark, which
gives it a Health Status of Good, though I wonder at what the columns
actually mean, in particular:

ID Attribute Name Current Worst Threshold Raw Values
05 Reallocated Sectors Count 100 100 50 All zeros
0A Spin Retry Count 253 100 30 All zeros

The full log is appended.

So next I ran MemTest on it - the PC has 4GB RAM and the CPU is an
Intel Core2 Duo running at 2.60GHz, and the test took about 3 hours 55
minutes to do a pass, which definitely seems very slow to me, but no
memory errors either.

However, I did notice that the two sticks were different makes, so may
not have been well matched, but, if that was an issue, why it had only
recently become so, I couldn't fathom. Nevertheless, as I still had 4GB
from P1 (see below) which I upgraded to 8GB, I swapped that in so that
the RAM modules are now guaranteed to be properly matched, but it's made
no difference.

> Of course, I could skip the shingles problem by going for an SSD, but
> have not really explored this up til now.  Experience and thoughts on
> that would be welcome too.

Yes, the general opinion does seem to be that this is the way to go.

The first thing though, now that CrystalMark has spoken somewhat
unexpectedly, is to find out WTF is actually making the PC so slow. An
obvious thing to look for is malware, but I don't it's likely to be
that, as it's *MUCH* slower than 2 other Dell Precision M6300s at even
beginning to load the GRUB menu or OS, which is why I felt certain that
the HD was most likely to be responsible. In the table below, the
problem PC is P2, and while it displays the Dell POST screen for about
the same time as the other two, there is then a long and variable pause
before it displays the GRUB menu, and thereafter the WinXP boot is also
slower (both GRUB menus timeout after 3s).

PC Time to: GRUB/OS Load Win Logon
P1 8GB @ 2.8GHz - Dual-boot Ubuntu 18 & W7: 0:12 0:24 later
P2 4GB @ 2.6GHz - Dual-boot Ubuntu 18 & XP: 0:31-1:00+ 0:19+ later
P3 4GB @ 2.2GHz - XP: 0:08 0:06 later

I've compared the setup options between P1 & P2, and they are the same
in all the things they can be, in particular both are set to Minimal
POST checks.

Anyone got any comments to make about this?

Appendix - CrystalMark log:
----------------------------------------------------------------------------
CrystalDiskInfo 8.17.14 (C) 2008-2022 hiyohiyo
Crystal Dew World:
https://crystalmark.info/
----------------------------------------------------------------------------

OS : Windows XP Professional SP3 [5.1 Build 2600] (x86)
Date : 2023/04/06 23:14:12

-- Controller Map ----------------------------------------------------------
- Ricoh SD/MMC Host Controller [ATA]
- Ricoh Memory Stick Controller [ATA]
- Ricoh xD-Picture Card Controller [ATA]
+ Intel(R) ICH8M Ultra ATA Storage Controllers - 2850 [ATA]
- Primary IDE Channel (0)
+ Intel(R) ICH8M 3 port Serial ATA Storage Controller - 2828 [ATA]
+ Primary IDE Channel (0)
- TOSHIBA MQ01ABD050V

-- Disk List ---------------------------------------------------------------
(01) TOSHIBA MQ01ABD050V : 500.1 GB [0/1/0, pd1]

----------------------------------------------------------------------------
(01) TOSHIBA MQ01ABD050V
----------------------------------------------------------------------------
Model : TOSHIBA MQ01ABD050V
Firmware : AX0G1Q
Serial Number : 43PDW013T
Disk Size : 500.1 GB (8.4/137.4/500.1/500.1)
Buffer Size : 16384 KB
Queue Depth : 32
# of Sectors : 976773168
Rotation Rate : 5400 RPM
Interface : Serial ATA
Major Version : ATA8-ACS
Minor Version : ----
Transfer Mode : SATA/300 | SATA/300
Power On Hours : 19576 hours
Power On Count : 63735 count
Temperature : 34 C (93 F)
Health Status : Good
Features : S.M.A.R.T., APM, NCQ, GPL
APM Level : 00FEh [OFF]
AAM Level : ----
Drive Letter : C: D:

-- S.M.A.R.T. --------------------------------------------------------------
ID Cur Wor Thr RawValues(6) Attribute Name
01 100 100 _50 000000000000 Read Error Rate
02 100 100 _50 000000000000 Throughput Performance
03 100 100 __1 000000000435 Spin-Up Time
04 100 100 __0 0000000101FA Start/Stop Count
05 100 100 _50 000000000000 Reallocated Sectors Count
07 100 100 _50 000000000000 Seek Error Rate
08 100 100 _50 000000000000 Seek Time Performance
09 _52 _52 __0 000000004C78 Power-On Hours
0A 253 100 _30 000000000000 Spin Retry Count
0C 100 100 __0 00000000F8F7 Power Cycle Count
BF 100 100 __0 00000000003A G-Sense Error Rate
C0 __1 __1 __0 00000000EB0F Power-off Retract Count
C1 _94 _94 __0 00000001022B Load/Unload Cycle Count
C2 100 100 __0 003300070022 Temperature
C4 100 100 __0 000000000000 Reallocation Event Count
C5 100 100 __0 000000000000 Current Pending Sector Count
C6 100 100 __0 000000000000 Uncorrectable Sector Count
C7 200 200 __0 000000000000 UltraDMA CRC Error Count
DC 100 100 __0 000000000000 Disk Shift
DE _52 _52 __0 000000004B85 Loaded Hours
DF 100 100 __0 000000000000 Load/Unload Retry Count
E0 100 100 __0 000000000000 Load Friction
E2 100 100 __0 0000000000B6 Load 'In'-time
F0 100 100 __1 000000000000 Head Flying Hours

-- IDENTIFY_DEVICE ---------------------------------------------------------
0 1 2 3 4 5 6 7 8 9
000: 0040 3FFF C837 0010 0000 0000 003F 0000 0000 0000
010: 2020 2020 2020 2020 2020 2034 3350 4457 3031 3354
020: 0000 8000 0000 4158 3047 3151 2020 544F 5348 4942
030: 4120 4D51 3031 4142 4430 3530 5620 2020 2020 2020
040: 2020 2020 2020 2020 2020 2020 2020 8010 0000 2F00
050: 4000 0200 0000 0007 3FFF 0010 003F FC10 00FB 0110
060: FFFF 0FFF 0007 0007 0003 0078 0078 0078 0078 0000
070: 0000 0000 0000 0000 0000 001F 0F06 0004 004C 0040
080: 01F8 0000 746B 7D69 6163 7469 BC41 6163 203F 0038
090: 0038 00FE FFFE 0000 0000 0000 0000 0000 0000 0000
100: 6030 3A38 0000 0000 0000 0000 6003 0000 5000 0394
110: B4A8 6F9A 0000 0000 0000 0000 0000 0000 0000 401C
120: 401C 0000 0000 0000 0000 0000 0000 0000 0029 0000
130: 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000
140: 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000
150: 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000
160: 0000 0000 0000 0000 0000 0000 0000 0000 0003 0000
170: 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000
180: 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000
190: 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000
200: 0000 0000 0000 0000 0000 0000 003D 0000 0000 4000
210: 0000 0000 0000 0000 0000 0000 0000 1518 0000 0000
220: 0000 0000 101F 0000 0000 0000 0000 0000 0000 0000
230: 0000 0000 0000 0000 0001 0080 0000 0000 0000 0000
240: 0000 0000 0000 0000 0000 0000 0000 0000 0000 0000
250: 0000 0000 0000 0000 0000 E1A5

-- SMART_READ_DATA ---------------------------------------------------------
+0 +1 +2 +3 +4 +5 +6 +7 +8 +9 +A +B +C +D +E +F
000: 10 00 01 0B 00 64 64 00 00 00 00 00 00 00 02 05
010: 00 64 64 00 00 00 00 00 00 00 03 27 00 64 64 35
020: 04 00 00 00 00 00 04 32 00 64 64 FA 01 01 00 00
030: 00 00 05 33 00 64 64 00 00 00 00 00 00 00 07 0B
040: 00 64 64 00 00 00 00 00 00 00 08 05 00 64 64 00
050: 00 00 00 00 00 00 09 32 00 34 34 78 4C 00 00 00
060: 00 00 0A 33 00 FD 64 00 00 00 00 00 00 00 0C 32
070: 00 64 64 F7 F8 00 00 00 00 00 BF 32 00 64 64 3A
080: 00 00 00 00 00 00 C0 32 00 01 01 0F EB 00 00 00
090: 00 00 C1 32 00 5E 5E 2B 02 01 00 00 00 00 C2 22
0A0: 00 64 64 22 00 07 00 33 00 00 C4 32 00 64 64 00
0B0: 00 00 00 00 00 00 C5 32 00 64 64 00 00 00 00 00
0C0: 00 00 C6 30 00 64 64 00 00 00 00 00 00 00 C7 32
0D0: 00 C8 C8 00 00 00 00 00 00 00 DC 02 00 64 64 00
0E0: 00 00 00 00 00 00 DE 32 00 34 34 85 4B 00 00 00
0F0: 00 00 DF 32 00 64 64 00 00 00 00 00 00 00 E0 22
100: 00 64 64 00 00 00 00 00 00 00 E2 26 00 64 64 B6
110: 00 00 00 00 00 00 F0 01 00 64 64 00 00 00 00 00
120: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
130: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
140: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
150: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
160: 00 00 00 00 00 00 00 00 00 00 82 00 78 00 00 5B
170: 03 00 01 00 02 78 00 00 00 00 00 00 00 00 00 00
180: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
190: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
1A0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
1B0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
1C0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
1D0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
1E0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
1F0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 52

-- SMART_READ_THRESHOLD ----------------------------------------------------
+0 +1 +2 +3 +4 +5 +6 +7 +8 +9 +A +B +C +D +E +F
000: 10 00 01 32 00 00 00 00 00 00 00 00 00 00 02 32
010: 00 00 00 00 00 00 00 00 00 00 03 01 00 00 00 00
020: 00 00 00 00 00 00 04 00 00 00 00 00 00 00 00 00
030: 00 00 05 32 00 00 00 00 00 00 00 00 00 00 07 32
040: 00 00 00 00 00 00 00 00 00 00 08 32 00 00 00 00
050: 00 00 00 00 00 00 09 00 00 00 00 00 00 00 00 00
060: 00 00 0A 1E 00 00 00 00 00 00 00 00 00 00 0C 00
070: 00 00 00 00 00 00 00 00 00 00 BF 00 00 00 00 00
080: 00 00 00 00 00 00 C0 00 00 00 00 00 00 00 00 00
090: 00 00 C1 00 00 00 00 00 00 00 00 00 00 00 C2 00
0A0: 00 00 00 00 00 00 00 00 00 00 C4 00 00 00 00 00
0B0: 00 00 00 00 00 00 C5 00 00 00 00 00 00 00 00 00
0C0: 00 00 C6 00 00 00 00 00 00 00 00 00 00 00 C7 00
0D0: 00 00 00 00 00 00 00 00 00 00 DC 00 00 00 00 00
0E0: 00 00 00 00 00 00 DE 00 00 00 00 00 00 00 00 00
0F0: 00 00 DF 00 00 00 00 00 00 00 00 00 00 00 E0 00
100: 00 00 00 00 00 00 00 00 00 00 E2 00 00 00 00 00
110: 00 00 00 00 00 00 F0 01 00 00 00 00 00 00 00 00
120: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
130: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
140: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
150: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
160: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
170: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
180: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
190: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
1A0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
1B0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
1C0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
1D0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
1E0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
1F0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 36

[Carlos E.R.]

On 2023-04-07 01:02, Martin Gregorie wrote:
> On Thu, 6 Apr 2023 23:22:03 +0200, Carlos E.R. wrote:
>
>> On 2023-04-06 17:02, Java Jive wrote:
>>> Please excuse the Linux/Windows crosspost, this is a question about HD
>>> hardware relevant to both OSs!
>>>
>>> I have a Dell Precision M6300 that is slowing down really badly, and I
>>> suspect, but have yet to prove, that the HD is failing.
>>
>> At least in Linux, this is easy to check. Assuming the drive is
>> /dev/sda, do, as root:
>>
>> smartctl -a /dev/sda
>>
>> Check these two lines:
>>
>> ID# ATTRIBUTE_NAME FLAG VALUE WORST THRESH TYPE UPDATED
>> WHEN_FAILED RAW_VALUE
>>
> I like to keep an eye on disk ages and potential problems, so I have a
> weekly cronjob that runs smartctl and emails me the report it produces.

You could simply run the daemon, smartd.


--
Cheers, Carlos.

[Java Jive]

On 07/04/2023 00:17, Java Jive wrote:
>
> PC                                 Time to:  GRUB/OS Load  Win Logon
> P1 8GB @ 2.8GHz - Dual-boot Ubuntu 18 & W7:  0:12          0:24 later
> P2 4GB @ 2.6GHz - Dual-boot Ubuntu 18 & XP:  0:31-1:00+    0:19+ later
> P3 4GB @ 2.2GHz - XP:                        0:08          0:06 later

Perhaps I should have explained that these are times for each PC to get
to the logon screen from a state of hibernation, in other words, from
no power being consumed through reloading the previous state to
displaying the logon screen.

[Carlos E.R.]

On 2023-04-07 01:17, Java Jive wrote:
> On 06/04/2023 16:02, Java Jive wrote:

...

I don't see anything wrong in this disk.

--
Cheers, Carlos.

[Davey]

On Thu, 6 Apr 2023 23:22:03 +0200

I tried this on my laptop, which has no performance problems that I
know of, and this is the result:
-------------------------------------------------------------------------------
SMART Attributes Data Structure revision number: 1
Vendor Specific SMART Attributes with Thresholds:

ID# ATTRIBUTE_NAME FLAG VALUE WORST THRESH TYPE UPDATED WHEN_FAILED
RAW_VALUE

1 Raw_Read_Error_Rate 0x0032 100 100 050 Old_age Always 0
5 Reallocated_Sector_Ct 0x0032 100 100 050 Old_age Always 0

(multiple similar lines), then finally:

245 Unknown_Attribute 0x0032 100 100 050 Old_age Always 101664

SMART Error Log Version: 1
No Errors Logged

SMART Self-test log structure revision number 1
Num Test_Description Status Remaining
LifeTime(hours) LBA_of_first_error # 1 Short offline Completed
without error 00% 993 -

Selective Self-tests/Logging not supported
-------------------------------------------------------------------------------

.. which does not match your '000' requirement for the last of the three
columns.

This is on a Linux system, BTW.
--
Davey

[Chris Green]

Jim Kelly <inv...@invalid.net> wrote:
> On 06/04/2023 17:11, J.O. Aho wrote:
> >
> > I don't recommend QVO as it wears out faster.
>
> How fast? 5 years, 4 years, 3 years or just 6 months?
>
> For most people if a hard disk lasts for 5 years then they have done
> very well indeed.

In 30 years or more of PC ownership I think I have only had one or two
disk drives fail, they mostly just get pensioned off when disk sizes
are such that the space the old drives have is not worth bothering
with.

For example the 2TB WD 'My Book' backup (two 1TB drives) which I
bought in 2009 and ran continuously until 2020 or thereabouts still
boots and makes its backup available if I need them.

--
Chris Green
·

[Java Jive]

Yes, yet the difference in time between P1 & P2 to reach the GRUB menu,
and the difference in speed with which it is drawn when it is reached -
P1 is almost instantaneous, while P2 painfully draws first the border
from the bottom up and then fills in the menu - is very, very
noticeable, and, as that is the moment when the HD is first accessed, I
felt sure that it must be the problem, but apparently not.

I now declare the problem officially a mystery!

[Paul]

On 4/6/2023 4:51 PM, J.O. Aho wrote:
> On 4/6/23 19:53, Jim Kelly wrote:
>> On 06/04/2023 17:11, J.O. Aho wrote:
>>>
>>> I don't recommend QVO as it wears out faster.
>>
>> How fast? 5 years, 4 years, 3 years or just 6 months?
>
> Time depends on how much you write to the disk, this includes the resizing of the paging file that windows does quite frequently in the background. You can look at the product warranty for the 1T evo 600 TBW (Max 5 years) vs 1T qvo 360 TBW (Max 3 years), that already hints that Samsung expects the qvo to have shorter lifespan than the evo, sure this number don't mean that all qvo will just last 3 years + 1 day or 360TB + 1B.
>

WinXP was like this. And the pagefile happened
to have just dreadful fragmentation, leading users to
be twiddling thumbs while it "un0wound". Just an awful design.

virtual memory
pagefile.sys

Win10/Win11 are like this:

virtual memory
Memory Compressor
pagefile.sys

The first discussions about pagefile and SSDs, happened when Sinovsky
was still around, in the Windows 7 era. Presumably, that's when an
internal project was underway, to modify how it works.

What you should see happen today, is only "transient"
consumption "hits" the pagefile. Most of the time, like
even when starting a single application, there is no
activity on the pagefile at all. If there isn't enough
memory to run an application, it just errors out. It
does not torture you by "paging out one of the other apps".

When the OS is given 1GB of memory, hard faults are pretty
close to zero, and the Memory Compressor hardly runs.
As you reduce memory to the OS, down to the 250-350MB range,
the hard fault rate shoots up, and you can find the Memory
Compressor railed on one core. And this behavior is "in defense"
of the pagefile, trying to avoid wasteful writes to it.

Modern Windows has lots of wasteful activity, that should
not be there, but paging is not it. There are ETL tracing
files, Search Indexer (wasteful!), Windows Defender (writes
while it scans!), Photos (is doing OCR of your photos and
saving the text in an SQLITE database!). It's a rampage of
"userland silliness" today, that is wearing your SSD.
There are as many as four potential services, that are
interested in "scraping for email addresses", but I have not
managed to catch them doing this. For the person who does not
seek to leave incriminating evidence on a computer, these new
OSes are your worst nightmare :-) Even if the info is not
exfiltrated, it just... should not be there.

Paul

[Paul]

https://ssd.userbenchmark.com/SpeedTest/3104/INTEL-SSDSC2BW240A4

Controller: LSI SandForce SF2281. <=== ding! ding! ding!

That's the SandForce data compressor, on writes.
Drive shoots up to 7W consumption spikes, during compression.

There aren't many compression methods, that can compress
in real time at 500MB/sec. And that's what the SandForce was doing.

And I thought only Kingston, used them.

SandForce were more common in the first generation, when there were
more of the smaller manufacturers (like maybe OCZ). But after that,
it was mostly Kingston that seemed to use them.

The other controllers don't do compression, or we'd have tales
of excess consumption for them too.

When you mentioned Intel, at first I thought the drive might
be Optane, but no, it's just Sandforce as root cause.

Paul

[Paul]

On 4/6/2023 7:46 PM, Java Jive wrote:
> On 07/04/2023 00:17, Java Jive wrote:
>>
>> PC                                 Time to:  GRUB/OS Load  Win Logon
>> P1 8GB @ 2.8GHz - Dual-boot Ubuntu 18 & W7:  0:12          0:24 later
>> P2 4GB @ 2.6GHz - Dual-boot Ubuntu 18 & XP:  0:31-1:00+    0:19+ later
>> P3 4GB @ 2.2GHz - XP:                        0:08          0:06 later
>
> Perhaps I should have explained that these are times for each PC to get to the logon screen from a state of hibernation, in other words, from no power being consumed through reloading the previous state to displaying the logon screen.
>

While in Windows, run HDTune benchmark, and look for "bad spots" in the curve.

Even when "Reallocated" is zero, the benchmark curve helps give the user
a "pre-warning" about surface wear on the platter.

In Linux, Gnome-disks has a benchmarking curve, but it isn't as well-developed
as some of the Windows third-party products (Gnome-disks needs to use more samples
and spend more time benching). Gnome-disks also allowed *write* benching at one time,
so be careful to not erase a disk by accident doing that.

SMART Reallocated works best, when defects are uniformly spread across the platter.

If all the spares are exhausted in a narrow swath of disk surface,
then Reallocated remains at zero, yet, the drive is in trouble.

It is for this reason, we bench with HDTune, as a pre-warning of trouble.
Disk health, is a combo of a clean benchmark run, as well as Reallocated is zero.
(The free benchmarker, only does read benches, so is "erasure-safe". It won't
erase a drive when you use it.)

https://www.hdtune.com/files/hdtune_255.exe # ten year old free version

In this example, the band at 53-54% is looking like a bad patch. I have a drive
here, that was used for WinXP, that it had a bad patch from 50% to 60% of
the disk and the performance was that low. Used to take forever to boot.
I replaced the drive, purely from a performance point of view. The SMART
was not really warning about this. But the bench "does not lie" :-) if
there is a localized surface problem, a bench can show you the problem.
(This isn't the absolute best benchmark curve, and sometimes it is
OS interference doing this. On a modern OS, you may need some uptime
to pass, before running your bench.)

https://images.sftcdn.net/images/t_app-cover-l,f_auto/p/315203d6-96d3-11e6-9cd9-00163ed833e7/3484776723/hd-tune-screenshot.jpg

On a hard drive, the outer circumference offers better rates than the
hub does, which is why the benchmark curve gently declines to half-rate.
When you see stairsteps in the bench curve, that is "zoned recording",
and the formatting of the tracks changes from one part of the disk
to another, on purpose. The stair steps then, are normal, and part of
design.

*******

Since you're using WinXP, the alignment of the drive could be wrong.
This may need to be re-aligned, during cloning.

Drive type Issue
512n No problem # This is the best drive type 512 byte external, 512 byte internal
512e Internal align # Most drives are 512 external, 4K internal. WinXP = realign please
SSD Internal align # 63 sector "tracks" do not align to power-of-two NAND storage. Realign it.

Windows 7 partitions are on 1 megabyte boundaries. This would be good with 512e HDD
or with an SSD.

WinXP partitions are related to CHS and Sectors = 63. Lots of numbers in
the WinXP metadata, are divisible by 63. This is not a good choice,
if the hard drive uses 4K internal sectors, and needs to do fractional
operations when handling clusters which don't align with the disk.

It's possible the original drive is a 512n, just based on "production era".

Linux fdisk should be able to tell you whether a drive is 512n or 512e.

On Western Digital, 512n drives were available on either WD Black or WD Gold.
Most others were 512e.

Summary: If I had to gamble on a hard drive, I would buy the smallest
WD Black I could find (WD5000LPSX 500GB), and hope for the best :-)
Hardly any technical information is available these days, for hard drives.
For that matter, I did not see a lot of *choice* for hard drives
for laptops -- SSD might be your only choice now. The 15mm thick hard
drives won't fit in a laptop. SSD are 7mm, laptop drives are 7mm or 9.5mm.
"Expansion" 2.5" external drives are generally the 15mm type. Too thick.

The stock of hard drives at my computer store, is very low. I can get
1TB 3.5" drives. They have no high capacity hard drives (how are you supposed
to do backups???). I would have to go online, to find more selection.
The retailers are not helping matters.

Paul

[David W. Hodgins]

Thanks. Finally an explanation that makes sense!

My first ssd drive is an OCZ, and it's still working fine after 10 years.

I have four ssd drives in this system.
Model=OCZ-AGILITY4, FwRev=1.5.2, SerialNo=OCZ-N82WMCWEEW3L4H5T
Model=KINGSTON SEDC400S37960G, FwRev=SAFM32.I, SerialNo=50026B727502FD10
Model=INTEL SSDSC2BW240A4, FwRev=DC12, SerialNo=BTDA329505PE2403GN
Model=KINGSTON SEDC400S37960G, FwRev=SAFM32.I, SerialNo=50026B727502FD90

For the OCZ AGILITY smartctl shows ...

ID# ATTRIBUTE_NAME FLAG VALUE WORST THRESH TYPE UPDATED WHEN_FAILED RAW_VALUE

1 Raw_Read_Error_Rate 0x0000 005 000 000 Old_age Offline - 5
3 Spin_Up_Time 0x0000 100 100 000 Old_age Offline - 0
4 Start_Stop_Count 0x0000 100 100 000 Old_age Offline - 0
5 Reallocated_Sector_Ct 0x0000 100 100 000 Old_age Offline - 1
9 Power_On_Hours 0x0000 100 100 000 Old_age Offline - 86564
12 Power_Cycle_Count 0x0000 100 100 000 Old_age Offline - 481
232 Lifetime_Writes 0x0000 100 100 000 Old_age Offline - 90965437028
233 Media_Wearout_Indicator 0x0000 092 000 000 Old_age Offline - 92

I've seen many people complaining about OCZ, but my experience with it has
been great. While I've had several spinning rust drives fail over the years,
I've yet to have an ssd drive fail.

I still have one spinning rust drive in this system, but almost never use it.
I do make multiple levels of backup.

Regards, Dave Hodgins

[Ken Blake]

On Fri, 7 Apr 2023 08:19:48 +0100, Chris Green <c...@isbd.net> wrote:

>Jim Kelly <inv...@invalid.net> wrote:
>> On 06/04/2023 17:11, J.O. Aho wrote:
>> >
>> > I don't recommend QVO as it wears out faster.
>>
>> How fast? 5 years, 4 years, 3 years or just 6 months?
>>
>> For most people if a hard disk lasts for 5 years then they have done
>> very well indeed.
>
>In 30 years or more of PC ownership

37 years for me.

> I think I have only had one or two
>disk drives fail,

Zero failures for me.

>they mostly just get pensioned off when disk sizes
>are such that the space the old drives have is not worth bothering
>with.

Same here.

[Vincent Coen]

Hello Java!

Friday April 07 2023 12:07, Java Jive wrote to All:


> Yes, yet the difference in time between P1 & P2 to reach the GRUB
> menu, and the difference in speed with which it is drawn when it is
> reached - P1 is almost instantaneous, while P2 painfully draws first
> the border from the bottom up and then fills in the menu - is very,
> very noticeable, and, as that is the moment when the HD is first
> accessed, I felt sure that it must be the problem, but apparently not.

It is no point really in looking at the Smart stats to see why boot time is
slow however, your drive is a 5400 RPM unit and I suspect the interface is
equally slow so that your real problem.

Replace it with say a WD Black running from 7200, 10,000 or even a 15k RPM
and you will see speed seriously improved BUT if you connection is below a
Sata 3 them improvement will not be that great but will be better.

A better solution assuming you again have a sata3 link is to use a SSD from
a good brand such as Samsung using the EVO range but check the performance
for Read and write (and this is slower than Reading).

Expected boot up time say on a Linux box is under 10 seconds, Windows
slower but that's Windows for you.
Note that for many Linux Distro's the system performs a check for
differences of hardware during the boot up sequence and some times this can
be a wee bit slow for which a solution is to turn off this test (unless you
have made a change to the hardware in the system).



Vincent

[Jim Lesurf]

In article <k9899h...@mid.individual.net>, J.O. Aho <us...@example.net>
wrote:

> I would recommend a SSD, no point in going for a HDD unless you need it
> for large scale storage 2TB+

I stopped buying spinning rust years ago. Had to look up what "shingled"
meant!... although I recognised the trick when I read the explanation.

Most recent drive I bought was an 8TB drive in a USB box+interface.
Blindingly fast with my newer machines. Use SSDs like this with my Linux
and RISC OS systems, no probs. Can't comment on Windows beyond saying
I've not bothered with it for c20 years. So I've trimmed the xposting to
omit doze. :-)

Jim

--
Please use the address on the audiomisc page if you wish to email me.
Electronics https://www.st-andrews.ac.uk/~www_pa/Scots_Guide/intro/electron.htm
biog http://jcgl.orpheusweb.co.uk/history/ups_and_downs.html
Audio Misc http://www.audiomisc.co.uk/index.html

[Carlos E.R.]

On 2023-04-07 14:46, Paul wrote:
> On a hard drive, the outer circumference offers better rates than the
> hub does, which is why the benchmark curve gently declines to half-rate.
> When you see stairsteps in the bench curve, that is "zoned recording",
> and the formatting of the tracks changes from one part of the disk
> to another, on purpose. The stair steps then, are normal, and part of
> design.

Once I did a brute force test of a new disk. I made a lot of partitions,
say 50. Then tested speed on each of them (probably using different
filesystems, too). I think I used "hdparm -tT /dev/sdXY", maybe some dd
read/write.

It turned out that the disk was significantly faster at about 1/3 of the
way. The centre was significantly slower.

Of course, the disk might be lying about its geometry.

--
Cheers, Carlos.

[Paul]

On 4/7/2023 10:49 AM, Ken Blake wrote:
> On Fri, 7 Apr 2023 08:19:48 +0100, Chris Green <c...@isbd.net> wrote:
>
>> Jim Kelly <inv...@invalid.net> wrote:
>>> On 06/04/2023 17:11, J.O. Aho wrote:
>>>>
>>>> I don't recommend QVO as it wears out faster.
>>>
>>> How fast? 5 years, 4 years, 3 years or just 6 months?
>>>
>>> For most people if a hard disk lasts for 5 years then they have done
>>> very well indeed.
>>
>> In 30 years or more of PC ownership
>
>
> 37 years for me.
>
>
>> I think I have only had one or two
>> disk drives fail,
>
>
> Zero failures for me.
>

Three in-service HDD failures. (Two Maxtor 40GB, a Barracuda 32550)
One infant-mortality (WD Black 1TB back to store for refund,
no motor operation brand new, six months ago).

Many drives "retired" before failure (rather than
being beaten on performance by newer drives). All
my crusty Seagate 500GB drives, are retired.

One 500GB drive has 57000 hours on it, and is in mint condition
as far as the benchmark curve is concerned. The drive does not
unload the heads while in service, either. It's been "flying"
for 57000 hours. That drive was retired, when the motherboard
failed. All the drives on the platform were cloned over, before
retirement.

Maybe two thirds of the HDD fleet is in good condition
(Reallocated == 0, benchmark curve Good).

*******

SSDs all good, but not used regularly. A recent upgrade
changed that (daily drivers are all SSD now). Ten SSDs total,
three in service. The other seven are for experiments.

One SSD was taken back to the store. Corsair Neutron. Some
SSD drives, when they are brand new, they need to be "written
from end to end", to freshen up the cells. Then, when you
bench the drive, it's a flat line at 450MB/sec on a SATA one.
On the Neutron, I was getting a little better than 100MB/sec
over most of the drive surface.

And I decided to take it back to the store "on principle".
That the drives should not sit on the shelf so long, that they
give crappy performance like that. If the computer store wants
to fix them, have at it kids.

I gave the store a print of my benches, so they could see
what the complaint is. The Corsair was originally an MLC drive,
which disappeared from the market when MLC NAND was no longer
available, and the same model came back with TLC chips in it instead
(and that means brand new firmware). And I don't know if a
different controller was used or not. But anyway, it went back
to the store, to keep the other SSDs company. Generally, you
do not see poor performance out of the box, with MLC based drives.

TLC or QLC drives, might need to be written from end to end,
to make the benches look "normal". Your choice of course.
If you're a trusting soul, the performance will "perk up"
a bit, as you use them. Can you wait that long ? :-) If
you are using a cheap drive, you'll never know until later,
whether 100MB/sec was actually "normal".

Paul

[Java Jive]

On 07/04/2023 00:17, Java Jive wrote:
>

> On 06/04/2023 16:02, Java Jive wrote:
>>
>> Please excuse the Linux/Windows crosspost, this is a question about HD
>> hardware relevant to both OSs!
>
> Thanks for all the replies, all of which I've read and noted.
>
>> I have a Dell Precision M6300 that is slowing down really badly, and I
>> suspect, but have yet to prove, that the HD is failing.
>
> This afternoon I got around to testing the HD with CrystalMark, which
> gives it a Health Status of Good, though I wonder at what the columns
> actually mean, in particular:
>
> ID  Attribute Name             Current  Worst  Threshold  Raw Values
> 05  Reallocated Sectors Count  100      100    50         All zeros
> 0A  Spin Retry Count           253      100    30         All zeros
>
> The full log is appended.

It's not the disk. I tried swapping in P3's disk and the long delay
between the POST screen and beginning to load anything, XP in that case,
was still there.

The other test I've done today is run Dell's own diagnostics on it,
which took a very long time. Like CrystalMark, no fault was found with
the disk. The only thing thrown up that might be significant is that
the CPU fan speed is not being detected. The fan spins up alright, but
the system board cannot sense its speed, so perhaps the system is
slowing down the CPU to keep things cool?

At any rate, given the disk is fine, I see no particular need to replace
it yet, and I'll probably have to leave this now until I have less work
on, but thanks for all the helpful comments.

Also, I'm still interested in SDDs for other reasons, probably a 1 &/or
2TB. Two particular models have been discussed, a Samsung 860 EVO 1TB &
the QVO equivalent. Any others that have given people good service?

[Paul]

If you schedule a CHKDSK on C: , does the drive pass at that ?

After the CHKDSK is done, you can try defragmenting it.

If you suspect the drive has issues, you can do a backup before
attempting the defrag.

The WinXP defragmenter, was written by
"Presidents Software" and is a third-party defragmenter, and
attempts to do "perfect defragmentation", placing files
shoulder-to-shoulder. But like all things software, there
are things it cannot move, and it can get into a snit while
it is working, so the results are not always perfect. But it
does a better job on defrag, than the Win7 or Win10 build-in
defragmenter (written by Microsoft). The difference between them,
is the Presidents Software defrag can take over eight hours to run,
while the Win7 and Win10 would take around ten minutes for their
run. A pretty stark difference.

Generally, click the drive icon, Properties : Tools : Defragment,
to access the defragmenter.

The defragmenter also has some option to automatically operate.
This could be related to prefetch items, but I don't recollect the details.

One day, I can hear a "tone" coming from the drive. I use ProcMon to
check, and the WinXP defragmenter is moving *one* sector, from and to, the
same physical location on the drive. The defragmenter is in a loop!
That's an example of when you "don't want an SSD" :-) I'm sure the
SSD would just love a bug like that. You don't have to move too much
data around on the drive, and then it does not stay in that loop. The
beauty of the hard drive (at the time), is the tone (while weak)
did give me a warning of weirdness incoming.

One other thing to check out, is SuperFetch (AKA "sysmain" service),
is a bit of a pig. Sometimes it is responsible for disk activity
unrelated to things users are interested in. I don't have a proposed
"quick fix" for you though :-) On WinXP, this might be where the .pf
files come from (Prefetch). You might have to look up what the
equivalent names are for that stuff, on Windows XP.

*******

Samsung is having a bad quarter, from a sales perspective,
so if you're to get a "deal" from them on an SSD, this is
about the right time for it. They will have to cut production
of chips, to eliminate the excess and reduce inventory level.
And once they do that, they won't have to drop the retail
price quite as much.

This will give you some idea, what is happening to the price.
Be careful who you buy from, to get the right price.

https://ca.camelcamelcamel.com/product/B08QB93S6R

Unless you really need extra space, I would not get the QVO.
A QVO might make a good data-only drive, but for an OS drive,
like for my daily driver, I'd get something better. it all
depends on whether you have good backup automation, as to how
much sense a QVO makes.

Since WinXP does not support GPT, you're kinda stuck at the 2TB
point. If you bought an even larger drive, it can still be
partitioned up to 2.2TB, and the free pool of the drive can still
take advantage internally, from the extra NAND flash. You still get
the full wear life of the drive, in units of TBW. The NAND usage
is unaffected by the partitioning details on the outside of the drive.
Sector 0 is not stored at location 0. There is an internal translation
table, that converts external storage location, to internal NAND block.
Wear leveling ensures all blocks get equal usage.

There are mechanisms for using all 4TB of a 4TB drive, on Windows XP.
Acronis Capacity Manager is one, and there was a separate driver
written by someone else, which does similar. The problem is, there is
no matching capability on Linux (so it's not seamless, across platform).
I did set up a manual mount command (loopback mount with offset),
to mount a partition above 2.2TB, it worked, but for some reason,
it was dog-slow (10MB/sec). And I backed out that setup and moved on.
It was fun to get working, but not a long term solution. You don't
need a partition table entry, to "mount" a random starting address of
a hard drive. As long as you know what that address is, and, that the
partition does not "cross" any other partitions, you can use it.

Paul

[Paul]

You would need some kind of disk-trace, to understand
exactly what that test is doing.

It could be, that the utility is intended to work with
a partition that is the same size as the entire disk.

To measure "stroke", needs an origin and a destination.

Perhaps the hdparm source code, has the answer ?

Paul

[Andy Burns]

Java Jive wrote:

> The only thing thrown up that might be significant is that the CPU fan
> speed is not being detected.  The fan spins up alright,

I was going to ask id the vents are full of fluff, but presumably you'd
have noticed if you can see the fan spinning?

[Carlos E.R.]

On 2023-04-07 23:38, Java Jive wrote:
> On 07/04/2023 00:17, Java Jive wrote:
>> On 06/04/2023 16:02, Java Jive wrote:


>
> It's not the disk.  I tried swapping in P3's disk and the long delay
> between the POST screen and beginning to load anything, XP in that case,
> was still there.
>
> The other test I've done today is run Dell's own diagnostics on it,
> which took a very long time.  Like CrystalMark, no fault was found with
> the disk.  The only thing thrown up that might be significant is that
> the CPU fan speed is not being detected.  The fan spins up alright, but
> the system board cannot sense its speed, so perhaps the system is
> slowing down the CPU to keep things cool?

That would be weird. A computer can indeed slow the CPU if it is hot,
but the fan speed should not matter.

Indeed, I have noticed in seemingly powerful, fanless laptopts and mini
pcs, that they start a a heavy cpu task at full speed, and after half a
minute they slow down because the CPU can not evacuate the heat at that
pace.

The machines are good for desktop use, where they respond fast to user
actions on a document, but not if the task is long.

Which is fine, if you know that design choice.


--
Cheers, Carlos.

[Paul]

On 4/7/2023 5:38 PM, Java Jive wrote:

> The only thing thrown up that might be significant is that the CPU fan speed is not being detected.

Fan speed is typically done with a three wire fan, and
the third wire has the two-pulse-per-rotation signal on it.

In the superIO, the hardware monitor uses a "period counter"
to count ticks between pulses. If the "period counter"
does a carry out (overflows), then the software concludes the
fan is not running or the RPM are too low. This puts a lower
limit on detected RPM. I had problems with this, on some
of my earlier desktops, they refused to put large enough
counter chains to do a good job on fan RPM. Maybe I would
get 1200 RPM, but 1199 RPM would register as 0 in the software.

Counter value Diagnostic Result

0 No pulses! Report zero RPM
1-255 Working Take inverse of period, in ticks
>255 Overflow Report zero, but fan RPM is actually < RPM_min limit of SuperIO.
Fan is spinning, but SW makes no distinction for "1199 RPM" case.

The fan is a 12V device. Either a 12V signal or a 5V signal
could come out of the fan. The SuperIO can have a circuit with
a couple resistors and a zener. The zener clips the fan
signal to a safe input level (for 5V TTL on the SuperIO). One
of the resistors, allows the open collector RPM signal to swing
to 12V if it wants (for 12V fans).

Paul

[Java Jive]

More a question of hearing, but there's no significant amount of fluff
blocking the vents.

Also, after posting last night, I installed SpeedFan and RMSpy on the
problem machine. SpeedFan shows that the temperatures are very
reasonable, mostly around 50-60C. RMSpy shows varying CPU speeds mostly
less then 1 GHz when idle, occasionally exceeding that number when
busier, for example while loading PaintShopPro.

[Paul]

Intel CPUs do Turbo for 28 seconds or 56 seconds.
After that time, they run at a lower speed. Some review
articles describe such policies.

What you're hearing, could be the Turbo profile.

Paul

[Carlos E.R.]

Maybe.

It is on Linux. One machine has a Intel(R) Pentium(R) CPU N3710 @
1.60GHz, another has a Intel(R) Celeron(R) CPU N3060 @ 1.60GHz. Yet
another is from another person.

--
Cheers, Carlos.

[Java Jive]

On 08/04/2023 13:22, Carlos E.R. wrote:
>
> On 2023-04-07 23:38, Java Jive wrote:
>>
>> The other test I've done today is run Dell's own diagnostics on it,
>> which took a very long time.  Like CrystalMark, no fault was found
>> with the disk.  The only thing thrown up that might be significant is
>> that the CPU fan speed is not being detected.  The fan spins up
>> alright, but the system board cannot sense its speed, so perhaps the
>> system is slowing down the CPU to keep things cool?
>
> That would be weird. A computer can indeed slow the CPU if it is hot,
> but the fan speed should not matter.

Yes, fair point, I hadn't quite thought that through.

[Java Jive]

On 08/04/2023 13:36, Paul wrote:
>
> On 4/7/2023 5:38 PM, Java Jive wrote:
>>
>> The only thing thrown up that might be significant is that the CPU fan
>> speed is not being detected.
>
> Fan speed is typically done with a three wire fan, and
> the third wire has the two-pulse-per-rotation signal on it.

Yes. As you say, the wiring is red, yellow, black to a standard mobo
connector. I presume the yellow is the fan-speed signal.

All of these being quite old laptops, the fans - there are two, 1 CPU,
1 GPU - tend to be noisy, and, unfortunately, they're a pain to swap,
you have to remove the keyboard, screen, and palm rest to get at them.
It's ridiculous really, considering they're a moving part so their
eventual failure is entirely predictable, why not mount them in
something equivalent to a drive bay so that they can be changed by
removing a couple of screws and sliding out a module, just like you
change a HD? It would add bugger-all to the price, might even reduce
it, because fans would then become more standardised, instead of the
proliferation of ever-so-slightly different models in similar laptops
that happens at the moment.

[Martin Liddle]

On 07/04/2023 22:38, Java Jive wrote:
>
> Also, I'm still interested in SDDs for other reasons, probably a 1 &/or
> 2TB.  Two particular models have been discussed, a Samsung 860 EVO 1TB &
> the QVO equivalent.  Any others that have given people good service?
>

I use Samsung EVO for mission critical stuff but the rest of my my
computers have Crucial SSDs of various models. In the past I have had
several conventional hard drive failures but so far (touchwood) no
failures at all with SSDs.
--
Martin Liddle,
Staveley, Chesterfield, Derbyshire UK

[Java Jive]

On 08/04/2023 18:14, Martin Liddle wrote:
>
> On 07/04/2023 22:38, Java Jive wrote:
>>
>> Also, I'm still interested in SDDs for other reasons, probably a 1
>> &/or 2TB.  Two particular models have been discussed, a Samsung 860
>> EVO 1TB & the QVO equivalent.  Any others that have given people good
>> service?
>
> I use Samsung EVO for mission critical stuff but the rest of my my
> computers have Crucial SSDs of various models.

Thanks, another vote for Samsung EVO then ...

> In the past I have had
> several conventional hard drive failures but so far (touchwood) no
> failures at all with SSDs.

Yes, previously, I've rather been put off SSD drives, because ...

- I have a SanDisk 120GB which needed replacing while under
warranty, a hassle which involved me driving 60 miles or so to Inverness
to get to the nearest drop-off point in their return system, but TBF the
replacement is still working;
- Additionally I have had about 25 USB sticks, of which 3 died early;

... which between them give a combined failure rate of at least 15%,
which I would have guessed was higher than that for conventional HDs,
but now, trying to remember back systematically as best as I can over
about 3 to 4 decades, actually I recall 5 early failures in at least
about 25 HDs, or a maximum of around 20%, so for me SSDs certainly have
performed no worse, and most probably have performed better, than
conventional HDs, which I wouldn't have expected to be the case without
systematically trying to recall the details of the HDs that I've had.

[Zaidy036]

On 4/8/2023 1:14 PM, Martin Liddle wrote:
> On 07/04/2023 22:38, Java Jive wrote:
>>
>> Also, I'm still interested in SDDs for other reasons, probably a 1
>> &/or 2TB.  Two particular models have been discussed, a Samsung 860
>> EVO 1TB & the QVO equivalent.  Any others that have given people good
>> service?
>>
> I use Samsung EVO for mission critical stuff but the rest of my my
> computers have Crucial SSDs of various models.  In the past I have had
> several conventional hard drive failures but so far (touchwood) no
> failures at all with SSDs.

- Just to add another vote for Samsung:
My C: SSD 840 EVO 500GB ATA Device (SATA SSD) was installed 13 Jan 2010
and now shows "68 days" left but what "days" means is not well defined
and comes up as 68% life on other check apps. Now reduces at <1% per month.

It has always been on 24/7 except during vacation and started on Win 7
Home 32 bit and on Win 10 Pro 64 bit for the last year.

I have a Samsung 870 EVO 500GB, cloned to my C:, sitting on my desk
waiting as insurance .... bought on Amazon for only $42 delivered.

[Char Jackson]

On Fri, 7 Apr 2023 22:38:46 +0100, Java Jive <ja...@evij.com.invalid>
wrote:

>Also, I'm still interested in SDDs for other reasons, probably a 1 &/or
>2TB. Two particular models have been discussed, a Samsung 860 EVO 1TB &
>the QVO equivalent. Any others that have given people good service?

People have been talking about Samsung QVO and EVO, but I'd like to toss
out a recommendation for Samsung Pro.

This article is one of many that tries to describe the differences
between the 3 Samsung product lines.

https://www.partitionwizard.com/clone-disk/samsung-qvo-vs-evo.html

Samsung QVO vs EVO vs PRO: What's the Difference?

How to Interpret Samsung SSD Model: QVO vs EVO vs PRO

In general, Samsung SSDs are mainly divided into two categories:
enterprise-level SSDs and consumer-level SSDs. Enterprise SSDs focus on
data integrity, followed by capacity and performance, and finally cost.
Consumer SSDs first seek cost, followed by capacity and performance, and
finally data integrity.

In addition, consumer-level SSDs are also divided into two categories:
SSDs for retail customers and SSDs for OEM customers. QVO, EVO, and PRO
usually appear in retail customer SSD models in the form of a suffix. To
some extent, these suffixes indicate different technologies applied in
NAND flash of these SSDs.

As we all know, an SSD often uses NAND flash to store data persistently.
When the NAND flash is made via different technologies, the SSD storage,
performance, and lifespan will vary accordingly. In Samsung SSDs, PRO
indicates the SSD uses MLC, EVO indicates the SSD uses TLC, and QVO
indicates the SSD uses QLC.

MLC, short for Multi-Level Cell, means that one memory cell can store 2
bits of data.

TLC, short for Triple-Level Cell, means that one memory cell can store 3
bits of data.

QLC, short for Quad-Level Cell, means that one memory cell can store 4
bits of data.

Cost: PRO SSD is the most expensive, then the EVO SSD, and finally the
QVO SSD.

Performance: performance of Samsung PRO SSD is the best, then the EVO
SSD, and finally the QVO SSD.

Lifespan: the lifespan of Samsung PRO SSD is the longest, then the EVO
SSD, and finally the QVO SSD.

...there is no other difference in technology among Samsung PRO, EVO,
and QVO SSDs, apart from NAND flash memory. But Samsung QVO, EVO, and
PRO SSDs still vary in performance and warranty (you regard it as
lifespan).

Samsung PRO SSD: It is currently the company’s flagship SATA SSD.
With MLC technology, its speed and the endurance rating or TBW make it
stand out from Samsung 860 QVO vs EVO vs PRO comparison. Although its
warranty period is similar with the 860 EVO series, its TBW is doubled.
But it’s also the most expensive one as well. On Amazon, it starts at
$87.99 (for 256GB).
Samsung EVO SSD: It’s one of the most popular SSD series in the
market and offers similar or near the performance of the 860 PRO SSD
series, but at a more affordable price. On Amazon, it starts at $59.98
(for 250GB).
Samsung QVO SSD: It is Samsung’s first consumer-grade quad-level
cell (QLC) NAND drive and has the same sequential read and write speed
with the 860 EVO. But the 4KB random read and write speeds and TBW can't
match with those of 860 EVO series. Its only advantage is price,
starting at $109.99 (for 1TB) on Amazon.

In a word, if you want a large-capacity and cost-effective SSD, you can
buy 860 QVO. If you have no special demand, 860 EVO is sufficient. If
you need an SSD that be used under heavy load, I recommended you to buy
860 PRO.

(end article quote, but there's more at the link above)

[Java Jive]

On 07/04/2023 22:38, Java Jive wrote:

Well now, the plot ever thickens ...

Booting from a W98 DOS Mode USB-stick, it gets to the config.sys menu in
just 9 secs, just the same as the others!

WTF is going on here???!!!

[David W. Hodgins]

On Sat, 08 Apr 2023 18:14:23 -0400, Java Jive <ja...@evij.com.invalid> wrote:
> Well now, the plot ever thickens ...
>
> Booting from a W98 DOS Mode USB-stick, it gets to the config.sys menu in
> just 9 secs, just the same as the others!
>
> WTF is going on here???!!!

Different controllers. Perhaps there is a problem with the sata controller or
one of the devices connected to it (in terms of the device being slow to
initialize), or a barely working sata cable/connector.

Start by re-seating all of the sata connectors.

Regards, Dave Hodgins

[Paul]

On the Samsung consumer SATA side, they have stopped with the Pro SKUs.
Some of the older lines had Pro, and likely honest to goodness MLC. I think
I may have one or two of those here. As far as I know, the 870 SATA line,
the top SKU is 870 EVO.

The NVMe still have Pro, but what do I find ?

https://www.tomshardware.com/reviews/samsung-990-pro-ssd-review/2

"After the pSLC cache runs out, the 990 Pro maintains around 1.4 GBps
in a direct-to-TLC mode. This is slower than expected and, indeed, the
rest of the drives in the test will all eventually out-write it."

Samsung likes to pretend they are "challenged by the English language"
and it's all a merry mixup. I find all sorts of bullshit in adverts,
which may have been put there by the vendors. How can you
nail them down on stuff like this, when the mothership is so "wishy-washy"
with terminology of "MLC-like". Either it is MLC or it is not.
There's no excuse for shit like that.

As soon as people talk of "SLC-cache" or the pSLC term (pseudo SLC) made up
in that review, then it's just TLC or QLC under the covers. And the writes happen
in two phases (which is not good for wear life).

At the current time, there is an issue with 980 Pro and 990 Pro
requiring a firmware upgrade before you use them. Ask your retailer
for details. The firmware was sparing out good sectors, and the
life indicator was rocketing down when nothing was going on.
The new firmware does not reverse the sparing situation either.
That's why you must do the firmware update immediately
if buying one brand new, and not wanting to lose service life.

With OCZ, it was changing flash chips used, in the middle
of a production run. My Corsair Neutron was like that too, showing
up with TLC in it. With WD/Seagate, it was trying to pass off
shingled drives as suitable usage in a NAS. Storage is like a
series of dark alleys downtown. Carry a flashlight at all times,
or you'll get mugged.

Paul

[Java Jive]

I'd already removed the HD to try the P3 one, but nevertheless I took it
out again and examined the connectors, and there's nothing visibly wrong
with them. I can't examine the mobo connectors or the controller chip
without a major dismantling of the laptop, and I'm too busy to do that ATM.

[Paul]

On 4/10/2023 8:07 AM, Java Jive wrote:
> On 08/04/2023 23:53, David W. Hodgins wrote:
>>
>> On Sat, 08 Apr 2023 18:14:23 -0400, Java Jive <ja...@evij.com.invalid> wrote:
>>>
>>> Well now, the plot ever thickens ...
>>>
>>> Booting from a W98 DOS Mode USB-stick, it gets to the config.sys menu in
>>> just 9 secs, just the same as the others!
>>>
>>> WTF is going on here???!!!
>>
>> Different controllers. Perhaps there is a problem with the sata controller or one of the devices connected to it (in terms of the device being slow to initialize), or a barely working sata
>> cable/connector.
>>
>> Start by re-seating all of the sata connectors.
>
> I'd already removed the HD to try the P3 one, but nevertheless I took it out again and examined the connectors, and there's nothing visibly wrong with them.  I can't examine the mobo connectors or the controller chip without a major dismantling of the laptop, and I'm too busy to do that ATM.
>

There are CRC counters on either end of the SATA link.
On a CRC error, a transmission can be repeated
(details unknown to me).

If you kink a SATA cable, it can have a high error rate.
SATA cables should be treated (roughly) like optical cable.
Don't exceed the bend radius allowed for SATA cables.
(The SATA cables that arrive in the mail, with an elastic
wrapped around them multiple times, that just makes me "nuts"
when I see that :-/ DONT DO THAT. )

SMART has the counter for the disk drive end. If there
is an OS performance counter, I don't know where that is buried.

*******

You could have a logjam in the PCI Express or DMI tree
in theory, but I doubt anyone, even with a pathological
setup, has managed that. It's possible to oversubscribe,
from the Southbridge end of the system, and not have
enough bandwidth. Normal computer usage, is never heavy
enough to make such as an observable condition. the computer
continues to work, that is not a problem, but the speeds
may no longer be optimal.

*******

CPUs have two levels of throttling. See references to
Dell "Throttlegate", for which the PDF is pretty hard to find.
It's a 25 page document, where a user researched why his
computer was slow. A couple Dell laptops seemed to have an
overaggressive BIOS throttling thing, where once the
computer slowed down, it would not speed up when the
anomalous operating conditions were removed. Only a reboot
would attempt to resolve the matter.

You can run Live Media, do "TORAM=yes" on the boot line,
then run a bench in there, to see what a different operating
environment finds. By running with media stored in RAM,
this removed disks from the picture, so you can study "slowness"
without disks.

Paul

[Dave]

On 06/04/2023 16:02, Java Jive wrote:

> Can anyone point to a UK source of reliable, genuinely new, moderately
> priced non-shingled laptop drives from about 500GB to 1.5TB?

WD model WD10JUCT is available from various suppliers for about £60.
It's intended for CCTV, DVRs and similar uses where the volume of data
written is similar to the volume read. OK the ones I have are new-old
stock dated 2017 - 2019.
--
Dave

[Paul]

On 4/11/2023 2:05 AM, Charlie+ wrote:
> On Mon, 10 Apr 2023 08:57:10 -0400, Paul <nos...@needed.invalid> wrote
> as underneath :
>
> snip

>> If you kink a SATA cable, it can have a high error rate.
>> SATA cables should be treated (roughly) like optical cable.
>> Don't exceed the bend radius allowed for SATA cables.
>> (The SATA cables that arrive in the mail, with an elastic
>> wrapped around them multiple times, that just makes me "nuts"
>> when I see that :-/ DONT DO THAT. )
>>

> snip
>> Paul
>
> Thanks I didnt know that! Have seen SATA cables supplied oem with
> motherboards folded but not that tightly. I once pulled a failing flat
> red type cable to bits out of interest and found Alu. single strands
> crimped to contacts inside the moulded ends which was a surprise to me.
> C+
>

There has to be a pronounced kink in it,
for the error counter to see an issue.

When you kink the cable, it crushes that white
insulating material and changes the transmission
line impedance.

Paul

[Jim Lesurf]

In article <u136b4$2ile3$1...@dont-email.me>, Paul <nos...@needed.invalid>
wrote:

> There has to be a pronounced kink in it, for the error counter to see an
> issue.

> When you kink the cable, it crushes that white insulating material and
> changes the transmission line impedance.

The 'kink' be better modelled as an added shunt component as it is
localised. However if the cable folded so lengths are close together, the
problem may be cross-field coupling between them. The fields that carry the
signal energy are outside the metal of the wires.

[Java Jive]

On 10/04/2023 13:07, Java Jive wrote:
>
> On 08/04/2023 23:53, David W. Hodgins wrote:
>>
>> On Sat, 08 Apr 2023 18:14:23 -0400, Java Jive <ja...@evij.com.invalid>
>> wrote:
>>>
>>> Well now, the plot ever thickens ...
>>>
>>> Booting from a W98 DOS Mode USB-stick, it gets to the config.sys menu in
>>> just 9 secs, just the same as the others!
>>>
>>> WTF is going on here???!!!
>>
>> Different controllers. Perhaps there is a problem with the sata
>> controller or one of the devices connected to it (in terms of the
>> device being slow to initialize), or a barely working sata
>> cable/connector.
>>
>> Start by re-seating all of the sata connectors.
>
> I'd already removed the HD to try the P3 one, but nevertheless I took it
> out again and examined the connectors, and there's nothing visibly wrong
> with them.  I can't examine the mobo connectors or the controller chip
> without a major dismantling of the laptop, and I'm too busy to do that ATM.

After a test re-imaging of the slow PC, a Dell Precision of M6300
laptop, with an old XP SysPrep image, I noticed that the DVD-Writer
wasn't showing in Device Manager. I swapped it with one of the others,
and now both DVD-Writers are showing up on both PCs, and on the problem
PC the long delay between the POST screen and beginning to load the OS
has been removed, so it seems that delay was being caused by the CD-DVD
unit being second in the boot search order, between USB (1st) and HD
(3rd), but no actual device being found at boot time.

However, in normal use, it's still behaving erratically and is unstable,
even after the re-imaging; usual symptom is the first explorer process
- the one that runs the Desktop as opposed to subsequent ones that run
the File Manager - crashing. However, that wasn't happening, at least
not that I can definitely remember, previously, so that may be related
to reverting to quite an old image. Having concluded my testing with
this image, I shall put the original image back on to see what happens.

However, whether or not extra problems have been introduced by the old
image, it still seems treacle-like slow, so there's still some problem
with it.

[J. P. Gilliver]

In message <u11k8d$2933d$1...@dont-email.me> at Mon, 10 Apr 2023 19:24:12,
Dave <da...@cyw.uklinux.net> writes

I think the ones for TV purposes are "purple", in WD's colour scheme.
Whether they're good (or even overkill), or bad, for general PC use,
I've no idea. (I _suspect_ they're probably good on reliability,
possibly only middling on speed, at least for random access.)

I don't know if they're different when it comes to the magnetic
surfaces, or just the controllers.
--
J. P. Gilliver

[J. P. Gilliver]

In message <u0s8nj$1bera$1...@dont-email.me> at Sat, 8 Apr 2023 18:36:51,
Java Jive <ja...@evij.com.invalid> writes
[]

>Yes, previously, I've rather been put off SSD drives, because ...

[]

>... which between them give a combined failure rate of at least 15%,
>which I would have guessed was higher than that for conventional HDs,
>but now, trying to remember back systematically as best as I can over
>about 3 to 4 decades, actually I recall 5 early failures in at least
>about 25 HDs, or a maximum of around 20%, so for me SSDs certainly have
>performed no worse, and most probably have performed better, than
>conventional HDs, which I wouldn't have expected to be the case without
>systematically trying to recall the details of the HDs that I've had.
>

_My_ nervousness about SSD drives has been the _manner_ of failure - and
that's probably unfairly based on my experience with USB sticks: my
_feeling_ is that solid-state memory devices fail suddenly with no
warning, whereas spinning drives _tend_ to decline gradually. (Not
always I know: I had one where - I think - the head or heads
spot-welded, so obviously the drive suddenly stopped spinning! [It had
been in a laptop with a heating problem. After all the usual methods
failed, I actually opened it in a clean cabinet we had at work, which is
how I know what happened: I freed it, and got 95-98% of the data off,
though condemned it thereafter.]) But on the whole HDs give advanced
indication of failure: make funny noises, or - more often, I think - no
obvious indication (unless you keep running HDTune), just get slower and
slower as the ECC works harder. (I know someone whose XP - or might have
been '9x - machine was eventually taking a quarter hour to boot! It was
fine once it _had_ booted, unless you did something disc-intensive.)
Then there's the bit about SSDs having a write counter, and suddenly
becoming read-only when it passes a certain point - do they still do
that? - and one product line (Intel I think) which became a brick
(neither read _nor_ write, so you couldn't even rescue the data) at that
point.

Obviously, if you back up properly, none of this should matter, but …
(-:
--
J. P. Gilliver

[Paul]

SSD drives have three-core ARM processors. There is a whack of
firmware in there, doing maintenance and maintaining "power-safe"
operations (keeping a copy of the translation table). Quite frequently,
when your hand is off the mouse, that three core processor
is doing stuff. The LED does not flash, when the three core processor
is on a rant.

The consumer SSD drives run without using a SuperCap. That's what
"power-safe" means, immediate power failure does not endanger
the "critical data" content. Power failures are also recorded
in SMART, so if you've been mis-treating your SSD, there is
a counter pointing at your misdeed. (I have a SATA to USB
converter that causes the power-failure counter to increment!
Not a builder of confidence, when a flush() was already sent.
This should not be happening. PC SATA SSD operation works fine.)

Early Enterprise drives had a SuperCap and the drive ran off
SuperCap energy, once the primary power feed was observed to
have gone away. This takes some of the pressure off writing
"power-safe" firmware. The SuperCap does not work (necessarily)
at rail voltage, and may use a boost converter to power
circuits on a failure. It only has to run for a second or two.
A few bucks worth of SuperCap would be enough, rather than
one of the $100 ones you could weld with :-)

Some consumer SSD drives, if you examine the PCB, you can see
the pads for the SuperCap (no part installed). The boost chip,
inductor and other bumpf, are also depopulated in the bill of
materials. Shopping on Ebay for a Supercap, isn't enough.

Some SSD drives have DRAM cache, cheaper ones do not. You won't
really find any discussion threads, where there is "evidence
these things exist". Presumably such cache, helps with wear
life and write amplification, best case.

USB sticks ? It's lucky they even have bypass caps.

There is nothing of value in a USB stick. Yes, a dinky
microcontroller is in there. A few USB sticks are "featureful",
but we have to take the word of manufacturer tech support,
and they're not known for information reliability. Most
modern USB sticks, die well before the computed wear life.
Even with no wear leveling, I should be able to write 600
times, and if it fails after 8 writes (dd.exe ==> ISO file),
you have to wonder.

They're just not in the same class.

There is much room for improvement, on USB sticks.

There are a couple sticks with both static and dynamic wear leveling.
But we cannot take some manufacturer tech support dood word for
this, because it could be abject marketing. Only the engineering
department at such a company, would know for sure.

Paul

[Paul]

https://www.tomshardware.com/reviews/surveillance-hard-drive-performance,3831-6.html

The bottom chart has Access Time, and the Access Time on the Purple is slow.
In a non-threaded storage situation (PC desktop), they would likely suck.
They would suck like my Seagate 4TB 5900RPM drive sucked yesterday :-)
Man is that thing slow. Good sequential though. For making backups.

https://www.tomshardware.com/reviews/surveillance-hard-drive-performance,3831-5.html

When you buy the wrong drive, you can always pretend it was for backups.

Paul

[Paul]

On 5/3/2023 4:55 PM, J. P. Gilliver wrote:
> In message <u0p3bc$qnoo$1...@dont-email.me> at Fri, 7 Apr 2023 08:46:36, Paul <nos...@needed.invalid> writes
> []
>> While in Windows, run HDTune benchmark, and look for "bad spots" in the curve.
> []
>>   https://www.hdtune.com/files/hdtune_255.exe   # ten year old free version
> []
>> On a hard drive, the outer circumference offers better rates than the
>> hub does, which is why the benchmark curve gently declines to half-rate.
>> When you see stairsteps in the bench curve, that is "zoned recording",
>> and the formatting of the tracks changes from one part of the disk
>> to another, on purpose. The stair steps then, are normal, and part of
>> design.
> []
> I have (what I think is a) fairly conventional HD (500G, in a second-hand laptop I bought in January), which displays a fairly normal HDTune curve - flattish up to about 40%, then gentle rolloff to about half speed: but then hops up again to a high speed for the last 2% or so! It's consistent - two pairs of runs (I always run it twice) about 3 months apart. (The second run also shows the Access Time - the yellow "milky way" of spots - along the bottom.)
>
> I'll try to include the last pair but they may not attach or propagate.

Well, you certainly got your moneys-worth.

It's like something from a Cracker Jack box.

One of your plots, has HDD sequential transfer rate and SSD-like access times.

The piece on the end, *might* be consistent with a short-stroked
drive. A short-stroked drive only uses half of the platter (the
outer half) and the heads never touch the hub. On a short stroked
drive, the transfer curve starts at full rate, but at the end, it
has only declined to about 80% or so. The transfer speed is
mostly consistent over the storage surface.

I am lucky enough, to have acquired just one short stroke drive,
and there is absolutely no notation in the part number, indicating
my Cracker Jack either. I have three drives of that model, two
normal, one is short stroked.

Mine is a a WD 500GB 3.5" drive which is using a 1TB platter inside,
both surfaces are certified, and they only use the first 500GB because
I only paid for a 500GB drive. They make up a batch of 1TB drive,
some become 500GB drives, some stay as 1TB drives. This solves the
problem, of having no platters available any more, to make the
500GB drives.

But that translation, makes no sense. You would not "jump the heads to
the middle of the disk" for the last bit of certified storage on the
drive. What I'm saying is, if your drive was short stroked, the height
of the material on the right, is consistent with a short stroke drive.
But that's just a (weak) attempt to explain where the height would come
from.

But your Access Times of 0.3ms blows the whole charade. Something
like that might happen, via a user adding some sort of Samsung caching
software. Still pretty hard to justify or believe.

There are two anomalies, and I cannot adequately explain either of them.

I'm not saying it's Space Aliens that did it, but it's Space Aliens.

If you were running RAID, had a 500GB hard drive, a 16GB Robson cache,
maybe there would be some way to rig that. (Check and see if your
storage is being run by the Intel RST driver.) Like first, I have to dream
up some materials to make this work. But then the "behavior" part of
the observation, still makes no sense. A Robson cache, I don't think
the curves look like that, and the label for the upper left would
not read the way it does either. There would be an artifact of
the presence of RST for the drive name.

Kudos on your puzzle.

Paul

[J. P. Gilliver]

In message <u2uvs8$1gp0k$1...@dont-email.me> at Wed, 3 May 2023 20:56:40,
Paul <nos...@needed.invalid> writes

>On 5/3/2023 4:55 PM, J. P. Gilliver wrote:
[]
>> I have (what I think is a) fairly conventional HD (500G, in a
>>second-hand laptop I bought in January), which displays a fairly
>>normal HDTune curve - flattish up to about 40%, then gentle rolloff to
>>about half speed: but then hops up again to a high speed for the last
>>2% or so! It's consistent - two pairs of runs (I always run it twice)
>>about 3 months apart. (The second run also shows the Access Time - the
>>yellow "milky way" of spots - along the bottom.)
>> I'll try to include the last pair but they may not attach or
>>propagate.
>
>Well, you certainly got your moneys-worth.

I'm very pleased with the machine: I paid £80 for it - it's a 15+", with
the 500G and 4G, though I presume the reseller probably inserted those.
(It can take 8G, but I wanted 7-32.) It seems very responsive. The make
[of the laptop] is "stone" (yes, with a lower case S), which I'd never
heard of. The only bad point is this weird loss of connection (but not
for YouTube and Google!) after a few hours, but that wasn't there when I
first got it (in January IIRR), so is something I've done.

>
>It's like something from a Cracker Jack box.
>
>One of your plots, has HDD sequential transfer rate and SSD-like access times.

I nearly always do two runs one after the other, so I'm guessing there's
some sort of buffering - or similar - that the elderly HDTune isn't
aware of: the pair of runs I did also have the very low access time on
the second one, though not _quite_ as drastically so (a _few_ of the
yellow "milky way" dots are still where they "should" be, and it shows
1.4 ms rather than 0.3).

>
>The piece on the end, *might* be consistent with a short-stroked
>drive. A short-stroked drive only uses half of the platter (the
>outer half) and the heads never touch the hub. On a short stroked
>drive, the transfer curve starts at full rate, but at the end, it
>has only declined to about 80% or so. The transfer speed is
>mostly consistent over the storage surface.

But on mine, it does decline to about 50%, or would if the odd jaggy
wasn't there at the end! So I don't _think_ it's short-stroked.

>
>I am lucky enough, to have acquired just one short stroke drive,
>and there is absolutely no notation in the part number, indicating
>my Cracker Jack either. I have three drives of that model, two
>normal, one is short stroked.
>
>Mine is a a WD 500GB 3.5" drive which is using a 1TB platter inside,
>both surfaces are certified, and they only use the first 500GB because
>I only paid for a 500GB drive. They make up a batch of 1TB drive,
>some become 500GB drives, some stay as 1TB drives. This solves the
>problem, of having no platters available any more, to make the
>500GB drives.

I'd have thought they'd make all drives with 1TB platters into 1 TB
drives, with those sold as 500G not being so sold to satisfy a demand
for 500G, but because on testing they found a big fault in the second
half. Rather like - _many_ decades ago, before PCs I think - you used to
sometimes get two versions of memory chips (or it might even have been
EPROMs) of a given capacity, one where one of the "enable" lines had to
be high and one where it had to be low, and it was fairly obvious that
these were chips made to twice the capacity but had had a fault found on
testing in one half or the other. (Obviously for HDs, if the fault was
in the _first_ half they'd scrap them.)

>
>But that translation, makes no sense. You would not "jump the heads to
>the middle of the disk" for the last bit of certified storage on the
>drive. What I'm saying is, if your drive was short stroked, the height
>of the material on the right, is consistent with a short stroke drive.
>But that's just a (weak) attempt to explain where the height would come
>from.
>
>But your Access Times of 0.3ms blows the whole charade. Something
>like that might happen, via a user adding some sort of Samsung caching
>software. Still pretty hard to justify or believe.
>
>There are two anomalies, and I cannot adequately explain either of them.
>
>I'm not saying it's Space Aliens that did it, but it's Space Aliens.

As I say, I suspect some sort of caching (or similar) hardware, either
in the drive or the PC, that is fooling HDTune on the second of two
successive runs. (It takes about 7 minutes, with HDTune left at its
default settings for block size etcetera.)

>
>If you were running RAID, had a 500GB hard drive, a 16GB Robson cache,
>maybe there would be some way to rig that. (Check and see if your
>storage is being run by the Intel RST driver.) Like first, I have to dream
>up some materials to make this work. But then the "behavior" part of
>the observation, still makes no sense. A Robson cache, I don't think
>the curves look like that, and the label for the upper left would
>not read the way it does either. There would be an artifact of
>the presence of RST for the drive name.

Certainly not (knowingly!) running any sort of RAID. (I'm pretty sure
there's only one HD - certainly the flap on the bottom of the machine [I
checked when buying it that there was a suitable access flap - I didn't
want one of these machines where you have to take the whole bottom cover
off to get at e. g. HD or RAM - is the normal size.)
>
>Kudos on your puzzle.

I won't worry about it if it gives me no trouble! And I'm backing up now
I've got my external drive back (unfortunately it was loaned to someone,
to get some of my data back, when the connection funny happened, so I
don't have a backup from before that).
>
> Paul
>
>
John
--
J. P. Gilliver

[J. P. Gilliver]

In message <u2uu1h$1ghd1$1...@dont-email.me> at Wed, 3 May 2023 20:25:21,
Paul <nos...@needed.invalid> writes

>On 5/3/2023 5:17 PM, J. P. Gilliver wrote:
>> In message <u11k8d$2933d$1...@dont-email.me> at Mon, 10 Apr 2023
>>19:24:12, Dave <da...@cyw.uklinux.net> writes
>>> On 06/04/2023 16:02, Java Jive wrote:
>>>
>>>> Can anyone point to a UK source of reliable, genuinely new,
>>>>moderately  priced non-shingled laptop drives from about 500GB to 1.5TB?
>>>
>>> WD model WD10JUCT is available from various suppliers for about £60.
>>>It's intended for CCTV, DVRs and similar uses where the volume of
>>>data written is similar to the volume read. OK the ones I have are
>>>new-old stock dated 2017 - 2019.
>> I think the ones for TV purposes are "purple", in WD's colour
>>scheme. Whether they're good (or even overkill), or bad, for general
>>PC use, I've no idea. (I _suspect_ they're probably good on
>>reliability, possibly only middling on speed, at least for random access.)
>> I don't know if they're different when it comes to the magnetic
>>surfaces, or just the controllers.
>
>https://www.tomshardware.com/reviews/surveillance-hard-drive-performance
>,3831-6.html
>
>The bottom chart has Access Time, and the Access Time on the Purple is slow.
>In a non-threaded storage situation (PC desktop), they would likely suck.

For access time, yes. Whether they'd be more reliable, I don't know - as
I said, I don't know if the magnetic arrangements - whether shingled,
say, or types of magnetic material - are any different to
non-CCTV/"purple" drives, or whether it's just the controller. (Though
of course that may affect reliability anyway.)

>They would suck like my Seagate 4TB 5900RPM drive sucked yesterday :-)
>Man is that thing slow. Good sequential though. For making backups.
>
>https://www.tomshardware.com/reviews/surveillance-hard-drive-performance
>,3831-5.html

I see they're using HDTune 2.55 - that's the ancient free one we use!

>
>When you buy the wrong drive, you can always pretend it was for backups.

(-:
>
> Paul
--
J. P. Gilliver