flash replacing hard disks?

  • Thread starter Thread starter Yousuf Khan
  • Start date Start date
Perhaps you are both trying to get by with 17-year-old drives _and_
media? Both wear out, you know. A floppy drive is real cheap, and
the new ones work when using recently manufactured media. Honest.

Felg, floppys suck! ...always have. I keep a few around for flashing
BIOS, but that's *it*. I keep a copy of thatever is on the floppy on hard
disk, because floppys suck worse than hard disk.

USB sticks are far better, even though it's well known how much I "love"
USB.
 
I always include a floppy too but I'm getting fed-up with them; I also
expect floppy disks to start increasing in price fairly soon. The big
names all seem to have given up even marketing them and the last time I
went to CompUSA, they had only store brand.

They already are (both drives and media). No one wants that crap anymore.
There are a couple of solutions here: buy a USB floppy drive or use a
USB flash drive. All BIOSs I've seen in the last 2 years are capable of
booting off both - might need to specify some USB "legacy" option in
BIOS Setup... and set the drive boot order correctly... and have a way
to make a USB flash drive bootable.

I gotta figure that one out. Small USB sticks are *cheap* (hell, .5GB
sticks are $30 locally, without silly rebates). One just to boot for a
flash might be a good investment. Then I could trash the floppy drives.
BTW, IME the nForce3/4 chipsets do not need a SATA driver loaded to get
WinXP to install to the HDD and then boot off it - the nForce chipset
seems to be compatible enough to run with the generic IDE driver. You
still have to load the specific nForce driver later to get the
performance of course.

I've never had a problem installing W2K. My problem with XP isn't at all
technical.
 
snip

I heard it with respect to thinkpad drives. And it certainly describes
the way the drive in my t23 died.

Someone stole the name. The problem with the IBM drives, AFAIK, was
finding home. They unloaded the heads on power down but didn't detect
it so retried forever (on either end of the power cycle).
*CLUNK*-*CLUNK*-*CLUNK*.
 
One good usage of a big drive is to enable a dual-boot Linux/Windows
machine, with plenty of space for both.

Sure, but drives are so cheap it's easy to split 'em. ;-)

I have Win on a SATA drive and Linux on the pATA. ...only because that's
the only way it works. ...and the SATA Win drive no longer works boots
properly. :-(
 
Might help for BIOS flash. But when you try installing NT/2k/XP/2k3
on a SATA/SCSI drive you have no chance to substitute CD for a floppy.
The installation is looking for the driver in A:\, and no option to
redirect it to D:\ or E:\ or whatever. Will anybody ever come up with
a flash memory reader that connects through floppy cable? I've seen
floppy/universal flash card reader combos, but for flash reader being
operational it had to connect to USB. Agree that the floppy interface
is sloooooooow, but this is where A:\ connects. Or any option in BIOS
to map A:\ to USB interface?

Yeah most BIOSs I've seen will end up mapping the USB flash to C: but I did
come across one early on which the BIOS was happy to see as a LS-120 and it
mapped it as A:.
 
I was wondering that too so I pulled up some Spansion datasheets. They claim
100K[*] for their "mirror bit" technology and 1M[*] for the "floating-gate"
parts.

2 million for these:
http://www.simpletech.com/oem/ideflashdrive/index.php

What's inside?

While we have a few of these around, they're held together with rivets, so
I'm not going to take the chance of opening it.

As a side note, http://www.dpie.com/storage/at3550.html these drives claim
to be good to 8 million erase/write cycles using NAND E2PROMs along with a
military temperature range -55c to +125c. Datasheet is here:
http://www.diamondpoint.co.uk/datasheets/storage/at3550datasheet.pdf

I'll bet they're not cheap, though.
 
mygarbage2000 said:
Size issues... Speed issues... Limited number of writes...

The Memtech solid-state drives Trent pointed out above seem to resolve these
issues fairly well. Up to 60GB, Average access .1ms, 8M cycles, 8 year
warranty seem to nail those concerns down (though STR seems to be left out of
the spec). Shock rating of 2000Gs is rather impressive too. You forgot cost
in your list though. ;-)
So it will
take quite some while to see 15k rpm SCSI drives dumped in favor of
solid state ;-)

With a .1ms average access, these make 15K RPM SCSI look turtle-slow.
Not in the next few (probably quite more than a
few) generations of servers...

Is there a need for solid-state drives for servers?
 
The Memtech solid-state drives Trent pointed out above seem to resolve these
issues fairly well. Up to 60GB, Average access .1ms, 8M cycles, 8 year
warranty seem to nail those concerns down (though STR seems to be left out of
the spec). Shock rating of 2000Gs is rather impressive too. You forgot cost
in your list though. ;-)


With a .1ms average access, these make 15K RPM SCSI look turtle-slow.


Is there a need for solid-state drives for servers?

It's application-driven: for example, if you really want tps to scream, use an
SSD for the database "after-image log". This file captures every write to the
database - and is the commit point for releasing file locks. As tps systems
are intrinsically synchronous, until that after-image file is updated with a
committing transaction, the entire system comes to a screeching stop, with a
whole system full of users waiting for their next screen refresh.

An alternative is to stick a non-volatile, secure write cache in the path from
the tps to the after-image log disk, and run something like Legato's
"PrestoServe" on it. Writes get captured by the secure write cache and the
file locks are released at that point, instead of waiting for the same write
to land on a cylinder. Plus, writes can be gathered at the cache and then
blasted to the spindle in block writes, instead of lots of individual write
transactions. As tps transactions are usually quite tiny in size, this cuts
average latency *and* queue depths down quite effectively...and results in
much happier users...

/daytripper (been there, designed all that, and thousands of AlphaServers
shipped with the secure write cache and a licensed instance of PrestoServe)
 
keith said:
Felg, floppys suck! ...always have. I keep a few around for flashing
BIOS, but that's *it*. I keep a copy of thatever is on the floppy on hard
disk, because floppys suck worse than hard disk.

Keith, don't change the rules in the middle of the game! ;-)

This thread started by pointing out (accurately) that there are a lot
of things for which you need a real, honest, floppy disk. CDs, even
when bootable, really aren't acceptable substitutes in many cases.
Then came the complaint that "my old floppy doesn't work, and neither
does my old media". I simply pointed out that everything wears out
eventually, and that working floppy drives and media are both cheap
and readily available.

Keith, I never said that floppies were wonderful. They aren't.
They're just _needed_!
 
Perhaps you are both trying to get by with 17-year-old drives _and_
media? Both wear out, you know. A floppy drive is real cheap, and
the new ones work when using recently manufactured media. Honest.

I purchased some new media within the past 5 years, and most of my
attempts to try and use said media was within the first few years of
owning those disks. As for the drives, the first couple I tried were
indeed rather old, so I had to purchase a new one which would work.
However even the new media was failing at rather alarmingly high
rates. Sure, might have just been a bad batch of disks, but I'm
through playing around with them.

Given that blank CDs are now cheaper than floppy disks (less than half
the price), I'll gladly burn a CD, even if I only need a couple
hundred KB of space.
 
Might help for BIOS flash. But when you try installing NT/2k/XP/2k3
on a SATA/SCSI drive you have no chance to substitute CD for a floppy.

Actually the newer nForce3 and nForce4 chipsets, as well as Intel's
i9xx series of chipsets, do not require any special drivers to install
WinXP/Win2K3 at the least. I don't know about WinNT 4.0 (bleah!) or
Win2K, though I suspect they should work fine too. Load up the OS in
a legacy mode that emulates standard IDE then load your drivers within
Windows to get access to the performance features of the SATA drive.

As far as I know it's really only the first generation of SATA
controllers that REQUIRED special drivers before they would work at
all.
 
Keith, don't change the rules in the middle of the game! ;-)

No rules changes, honest injun!
This thread started by pointing out (accurately) that there are a lot
of things for which you need a real, honest, floppy disk. CDs, even
when bootable, really aren't acceptable substitutes in many cases.
Then came the complaint that "my old floppy doesn't work, and neither
does my old media". I simply pointed out that everything wears out
eventually, and that working floppy drives and media are both cheap
and readily available.

I've bought *new* media and used *new* drives, and they're still
unreliable as all outdoors. What was written one day couldn't be read the
next. I had a few boxes of floppys a couple of years ago. Perhaps one in
four could be read in the drive they were written on.
Keith, I never said that floppies were wonderful. They aren't. They're
just _needed_!

It's about time to find something "more needed". If I was convinced I
could boot from crappy USB, I'd rather buy a few $10 sticks for
emergencies. IMO, floppys simply can't be trusted.
 
Tony said:
Given that blank CDs are now cheaper than floppy disks (less than half
the price), I'll gladly burn a CD, even if I only need a couple
hundred KB of space.

In cases like that I tend to go with a CD-RW or DVDxRW, that way you
don't have to waste all of that space, you can add to it later, or just
erase and start new.

Having said that, I've already run into a few CD-RW's that have run out
of their rewritable life.

Yousuf Khan
 
in Message id: <[email protected]>:

I was wondering that too so I pulled up some Spansion datasheets. They claim
100K[*] for their "mirror bit" technology and 1M[*] for the "floating-gate"
parts.

2 million for these:
http://www.simpletech.com/oem/ideflashdrive/index.php

What's inside?

While we have a few of these around, they're held together with rivets, so
I'm not going to take the chance of opening it.

That's certainly understandable.
As a side note, http://www.dpie.com/storage/at3550.html these drives claim
to be good to 8 million erase/write cycles using NAND E2PROMs along with a
military temperature range -55c to +125c. Datasheet is here:
http://www.diamondpoint.co.uk/datasheets/storage/at3550datasheet.pdf

Interesting. "Active remap" seems to imply that they're using tricks to move
things around to make that 8M cycles. From the above:

Active Remap™ Data Reliability Feature Endurance
# erase/write endurance 8 million cycles
# read endurance unlimited
# data Integrity 10 years"

I'll bet they're not cheap, though.

I'll bet. I note that they're all 3.5" form factor. I'd think the real market
would be in the 2.5" form.
 
It's application-driven: for example, if you really want tps to scream, use an
SSD for the database "after-image log". This file captures every write to the
database - and is the commit point for releasing file locks. As tps systems
are intrinsically synchronous, until that after-image file is updated with a
committing transaction, the entire system comes to a screeching stop, with a
whole system full of users waiting for their next screen refresh.

An alternative is to stick a non-volatile, secure write cache in the path from
the tps to the after-image log disk, and run something like Legato's
"PrestoServe" on it. Writes get captured by the secure write cache and the
file locks are released at that point, instead of waiting for the same write
to land on a cylinder. Plus, writes can be gathered at the cache and then
blasted to the spindle in block writes, instead of lots of individual write
transactions. As tps transactions are usually quite tiny in size, this cuts
average latency *and* queue depths down quite effectively...and results in
much happier users...

Interesting. Is nonvolatile == flash? ... or is it battery backed up DRAM?
 
On Thu, 06 Oct 2005 01:19:48 GMT, "(e-mail address removed)"


Actually the newer nForce3 and nForce4 chipsets, as well as Intel's
i9xx series of chipsets, do not require any special drivers to install
WinXP/Win2K3 at the least. I don't know about WinNT 4.0 (bleah!) or
Win2K, though I suspect they should work fine too. Load up the OS in
a legacy mode that emulates standard IDE then load your drivers within
Windows to get access to the performance features of the SATA drive.

As far as I know it's really only the first generation of SATA
controllers that REQUIRED special drivers before they would work at
all.

VIA SATA controller (vt800 chipset) surely requires - checked with
both 2k and XP. Same for a few less common SCSI cards (not plain
vanilla Adaptec). Either NVDA and INTC did their homework better
(work with regular IDE driver) or MSFT decides which drivers are
included in Setup and which have to be loaded from A:\.
 
Interesting. Is nonvolatile == flash? ... or is it battery backed up DRAM?

Not flash, of course, considering the write-cycles-to-wear-out problem in a
caching application ;-)

It was ultra-low power Sony static rams, kept alive with an on-board 3v coin
cell, all mounted on a PCI card and plugged into an PCI slot. If a system
totally died and had to be replaced, you could literally pull the card from
the failed box and stick it in the replacement, and the file system manager
would scarf any unwritten data and commit it to the disks. With the coin cell,
data hold time was measured in multiple weeks, even at the highest capacity
configuration.

Which led to an unintended application: our bios and firmware coders would use
the card to migrate their code between systems (systems that couldn't boot to
any io device in their current state), and system reliability testers used
hundreds of these cards to beat the snot out of their IO systems (think of a
TurboLaser maxed out at 192 PCI slots, and how do you do worst-case testing?)
because any single instance could consume the peak throughput of any bus it
was plugged into.

And it all started as a "midnite project" :-)

cheers

/daytripper
 
Bitstring <[email protected]>, from the
wonderful person "[email protected] said:
VIA SATA controller (vt800 chipset) surely requires - checked with
both 2k and XP. Same for a few less common SCSI cards (not plain
vanilla Adaptec). Either NVDA and INTC did their homework better
(work with regular IDE driver) or MSFT decides which drivers are
included in Setup and which have to be loaded from A:\.

It may also depend on which version of XP you are trying to install.
IIRC (and I may not be) SATA support arrived working right somewhere
around SP1 .. I think vanilla XP required drivers from floppy, and maybe
still does (if anyone is still installing 'pre SP1' XP).
 
VIA SATA controller (vt800 chipset) surely requires - checked with
both 2k and XP. Same for a few less common SCSI cards (not plain
vanilla Adaptec). Either NVDA and INTC did their homework better
(work with regular IDE driver) or MSFT decides which drivers are
included in Setup and which have to be loaded from A:\.

The Intel and nVidia controllers have a fallback mode which emulates a
standard IDE controller on their second generation of SATA controllers
(the i865/i875 series of SATA controllers did require drivers as I
recall). I wouldn't be surprised if VIA has done this as well with
their current SATA controller.

Note that sometimes this might require setting some option in the BIOS
for it to work though.

Ohh, and as GSV mentioned there are also some differences depending on
what version/service pack of Windows you are installing. WinXP SP2
definitely has some differences from SP1 or the original XP in this
regard, though I can't remember all of the specifics.
 
Back
Top