AHCI and IDE

[Robin Bignall]

A while ago I applied the registry fixes for Win7 Ult to support AHCI
when the system had been installed with IDE. I had already loaded my
motherboard's AHCI drivers. It apparently worked but I didn't see any
major change in performance. I went back to IDE for some reason that I
can't remember. I now find TRIM for my SSD will only enable if I use
AHCI so I changed back again. Possibly, I can't tell, because going
from AHCI to IDE to AHCI in BIOS did not appear to load any drivers each
time.

Looking at Device Manager "Disk Drives", my SATA SSD and 2 SATA HDDS are
described in properties as "ATA.... SCSI Disk device".

In "IDE ATA/ATAPI Controllers" I have 3 entries for" ATA Channel 0", 3
entries for "ATA Channel 1" and 3 entries for "Standard ACHI 1.0 SATA
Controller".

In "Storage Controllers" I have an entry "Intel C600 Series Chipset SATA
AHCI Controller" and one for "Microsoft iSCSI Initiator"

I have a feeling that this is screwed up somehow. What's the way to go?
Disable or uninstall all the controllers and disks, reboot and let
Windows find the right drivers?

 

[RayLopez99]

I have a feeling that this is screwed up somehow. What's the way to go?

Disable or uninstall all the controllers and disks, reboot and let

Windows find the right drivers?

I've built many a PC system but am no guru, and with that in mind when you "Disable or uninstall all the controllers and disks, reboot" what exactly does that mean? From inside of Control Panel in Windows? Or something in the BIOS? I assume that you mean the former.

Looking forward to seeing whoever answers your question, as I am leaning towards buying an SSD but I see more and more complications with them.

RL

 

[Charlie Hoffpauir]

I've built many a PC system but am no guru, and with that in mind when you "Disable or uninstall all the controllers and disks, reboot" what exactly does that mean? From inside of Control Panel in Windows? Or something in the BIOS? I assume that you mean the former.

Looking forward to seeing whoever answers your question, as I am leaning towards buying an SSD but I see more and more complications with them.

RL

I'm partly in your corner.... I certainly want to learn more, but I've
been using a SSD as my C drive for over a year and have seen none of
the problems that Robin is seeing. My Windows is set up for AHCI
drives, but I've also set up RAID-0 for my data drives. Looking at
SMART for all the drives using Hard Disk Sentinel, it looks to me like
SMART is working and reporting. And the command line statement I
posted to the other thread tells me TRIM is enabled.... Maybe my
situation is "Ignorance is bliss".

 

[Robin Bignall]

I'm partly in your corner.... I certainly want to learn more, but I've
been using a SSD as my C drive for over a year and have seen none of
the problems that Robin is seeing. My Windows is set up for AHCI
drives, but I've also set up RAID-0 for my data drives. Looking at
SMART for all the drives using Hard Disk Sentinel, it looks to me like
SMART is working and reporting. And the command line statement I
posted to the other thread tells me TRIM is enabled.... Maybe my
situation is "Ignorance is bliss".

I see exactly the same with Sentinel (which I bought yesterday). Let me
reiterate strongly I HAVE HAD NO PROBLEMS WITH SSD. Everything works
fine.
I just wonder about things like:
Who, what and when does TRIM get run?
Why are my SATA disks called SCSI devices in Device Mgr?
Why do I have 3 sets of 3 drivers for SATA disks in DM - IDE/ATAPI
controllers?
Why an iSCSI initiator in DM - Storage Controllers - what does that
control?

Incidentally, in Sentinel, if I click on the SSD and Information, down
the list a long way is something like SSD Features. The final entry
says "OS TRIM Function...... Supported, Disabled"

My best guess at what this means, since TRIM IS enabled in Win7, is ...
WTF does it mean?

 

[RayLopez99]

My best guess at what this means, since TRIM IS enabled in Win7, is ...

WTF does it mean?

Well you may be happy with SSD, as you say, but how do you know it's working properly? It's like having a high performance car and hearing a tiny persistent tinny knocking sound...that would give me pause.

RL

 

[DevilsPGD]

In the last episode of <[email protected]>,

I'm partly in your corner.... I certainly want to learn more, but I've
been using a SSD as my C drive for over a year and have seen none of
the problems that Robin is seeing. My Windows is set up for AHCI
drives, but I've also set up RAID-0 for my data drives. Looking at
SMART for all the drives using Hard Disk Sentinel, it looks to me like
SMART is working and reporting. And the command line statement I
posted to the other thread tells me TRIM is enabled.... Maybe my
situation is "Ignorance is bliss".

Most SSDs have some sort of garbage collection that partially negates
the need for TRIM. It doesn't really solve the problem as well as having
the OS say "This block isn't needed anymore", but it helps.

I'm not fully clear about how exactly drives do their own garbage
collection either, as far as I can tell they either need a basic
understanding of filesystem data structures (which won't exist in a
RAID-0/5 environment), or they would need to rely on the OS occasionally
wiping the drive with NULLs (which is more or less how, in the virtual
machine world, a guest OS can indicate that certain blocks aren't needed
anymore)

 

[Charlie Hoffpauir]

I see exactly the same with Sentinel (which I bought yesterday). Let me
reiterate strongly I HAVE HAD NO PROBLEMS WITH SSD. Everything works
fine.
I just wonder about things like:
Who, what and when does TRIM get run?
Why are my SATA disks called SCSI devices in Device Mgr?
Why do I have 3 sets of 3 drivers for SATA disks in DM - IDE/ATAPI
controllers?
Why an iSCSI initiator in DM - Storage Controllers - what does that
control?

Incidentally, in Sentinel, if I click on the SSD and Information, down
the list a long way is something like SSD Features. The final entry
says "OS TRIM Function...... Supported, Disabled"

My best guess at what this means, since TRIM IS enabled in Win7, is ...
WTF does it mean?

I see the same thing.... but look at the box next to the statement.
The box is UNCHECKED, which in my perverse logic, means it IS enabled.

 

[Robin Bignall]

Well you may be happy with SSD, as you say, but how do you know it's working properly? It's like having a high performance car and hearing a tiny persistent tinny knocking sound...that would give me pause.

Look at performance in Sentinel of SSD versus an HDD. Speaks for
itself.

 

[Robin Bignall]

I see the same thing.... but look at the box next to the statement.
The box is UNCHECKED, which in my perverse logic, means it IS enabled.

That is a trim idea. I like it.

 

[Ting Hsu]

Well you may be happy with SSD, as you say, but how do you know it's working properly?  It's like having a high performance car and hearing a tinypersistent tinny knocking sound...that would give me pause.

I don't see what the problem is. Just put your OS and applications on
the SSD, and it either works just fine, or it doesn't work at all. If
there's this tiny persistent problem, then sure, it might fail in two
years, but until then, you've got this mighty fine sports car.

And when it does fail, just buy a new one and reload the OS and
applications onto it, from your original media (or online downloads).
All your data should have been saved away on spinning hard drives in a
raid 1 or a raid 5 set up, as big SSDs are really too expensive for
data storage, especially in a redundant raid set up. So upon a
failure, you've really only lost an afternoon - it's annoying, but far
from the end of the world.

 

[DevilsPGD]

In the last episode of

I don't see what the problem is. Just put your OS and applications on
the SSD, and it either works just fine, or it doesn't work at all. If
there's this tiny persistent problem, then sure, it might fail in two
years, but until then, you've got this mighty fine sports car.

The problem with SSDs is that after a few months your fine sports car
might only get the first 3 gears, 4-5 disappear without either TRIM or
garbage collection.

 

[RayLopez99]

Look at performance in Sentinel of SSD versus an HDD. Speaks for

itself.

Link please?

I did find the below, and it's not dispositive.

RL

http://www.technize.com/ssd-vs-hdd-comparison/

What? No mention of degradation over write repetitions for SSDs?
No write speed comparison?
Deceptive and inconclusive.
I’ll have to look for an article that will explain this properly now.

 

[RayLopez99]

All your data should have been saved away on spinning hard drives in a

raid 1 or a raid 5 set up, as big SSDs are really too expensive for

data storage, especially in a redundant raid set up. So upon a

failure, you've really only lost an afternoon - it's annoying, but far

from the end of the world.

I think this is sound advice. I just need confirmation on the fail rate of SSDs. For mechanical HDs it's about five years or longer.

RL

 

[Robin Bignall]

I think this is sound advice. I just need confirmation on the fail rateof SSDs. For mechanical HDs it's about five years or longer.

RL

Download a program called SSDLife and read how it works. It's been
monitoring my SSD since the day it was installed (August 12). Predicted
life at my current usage is Sept 7, 2021. This date has not changed in
4 months.

 

[DevilsPGD]

In the last episode of

I think this is sound advice. I just need confirmation on the fail rate
of SSDs. For mechanical HDs it's about five years or longer.

Like anything, it's luck. I've got a small stack of drives here that are
far older than that that work fine, but are too small to be useful or
have an outdated interface.

Right beside that stack is a stack of warranty-ready (under 3-year old)
drives that need to be wiped and sent back to be replaced.

I've got a lot more drives than a typical home user here, but in my
experience there's very little predictability as to how long drives will
last, or which will die first.

SSDs have been similar. I've had a couple failures, but most are going
strong after many moons, and some are starting to get replaced due to
capacity since 240GB+ SSDs are suddenly affordable.

 

[Loren Pechtel]

Download a program called SSDLife and read how it works. It's been
monitoring my SSD since the day it was installed (August 12). Predicted
life at my current usage is Sept 7, 2021. This date has not changed in
4 months.

I've seen the same pattern and it's obviously figuring some maximum
lifetime, it's *NOT* going on usage to get that. Extrapolating simply
based on the wear life would give a date far beyond that.

 

[Loren Pechtel]

In the last episode of


Like anything, it's luck. I've got a small stack of drives here that are
far older than that that work fine, but are too small to be useful or
have an outdated interface.

Right beside that stack is a stack of warranty-ready (under 3-year old)
drives that need to be wiped and sent back to be replaced.

I've seen the same problem but I think it's more a matter of the newer
stuff being junk.

Up until drives passed 500gb I never had a warranty issue. Since then
I've sent back several.

 

[DevilsPGD]

In the last episode of <[email protected]>,

I've seen the same problem but I think it's more a matter of the newer
stuff being junk.

Up until drives passed 500gb I never had a warranty issue. Since then
I've sent back several.

I've sent back everything from the 80GB range up to 1TB. I don't think
I've sent a 2TB back, and certainly not a 3TB, but I might simply not
have enough of them in service yet.

 

[RayLopez99]

Yes, that's consistent with Google's report...somewhere in my HD, see if I can find it...yes, copied and pasted below. 7% of HDs fail within a year, temperature has no real effect on failure (in fact, cold temperatures causeslightly more failures than hot), nor does how hard you use your drive have any effect. SMART only caught 56% of failures, 30-50% of HDs fail within 5 years.

Interesting stuff I'm sure you'll agree, and scientifically sampled.

RL

In the last episode of

Like anything, it's luck.

//

http://labs.google.com/papers/disk_failures.pdf

But for those of you who just want to know the conclusions, here are the really interesting points of the article:

1) Hard drives fail a lot more often when young than we might like to think.. Google found that 3% of all hard drives fail within 3 months. Scarier still? 7% of all hard drives fail within 1 year. This means you have a 1 in 14 chance of your hard drive failing within the first year. So, just because your computer is new does not mean you don’t have to back it up.

2) Hard drives frequently fail without any warning at all. This isn’t really news to me, but it’s nice to see statistics that confirm it. Google found that only 56% of the hard drives that failed showed warning signs of failure before they died. So there is a 44% chance you will have no warning at all before your hard drive goes, one day it works, the next day it poofs..

3) Use is not a factor in hard drive failure. Google found no corolation between how much a hard drive had been used, and when it failed. We often hear, “well, but I only use my hard drives on weekends for 15 minutes, so shouldn’t it last forever?” This proves that is not true. Hard drives decay with age, regardless of use.

4) Temperature is probably not a factor in hard drive failure. Google foundno corolation between hard drive failure and drive temperature, but the study does acknowledge that the sample set is problematic, because google data centers are all temperature controlled. Previous studies have found a connection between temperature and drive failure, with too cold being more of a factor than too hot.

5) Google found the five year failure rate of hard drives to be 39% with amargin of error of about 9%. This means after five years, somewhere between 30% and 48% of all hard drives have failed. This tracks pretty closely to what we see here in the shop, where 4-6 year old computers often have dead hard drives.

 

[Charlie Hoffpauir]

Yes, that's consistent with Google's report...somewhere in my HD, see if I can find it...yes, copied and pasted below. 7% of HDs fail within a year, temperature has no real effect on failure (in fact, cold temperatures cause slightly more failures than hot), nor does how hard you use your drive have any effect. SMART only caught 56% of failures, 30-50% of HDs fail within 5 years.

Interesting stuff I'm sure you'll agree, and scientifically sampled.

RL




//

http://labs.google.com/papers/disk_failures.pdf

But for those of you who just want to know the conclusions, here are the really interesting points of the article:

1) Hard drives fail a lot more often when young than we might like to think. Google found that 3% of all hard drives fail within 3 months. Scarier still? 7% of all hard drives fail within 1 year. This means you have a 1 in 14 chance of your hard drive failing within the first year. So, just because your computer is new does not mean you don’t have to back it up.

2) Hard drives frequently fail without any warning at all. This isn’t really news to me, but it’s nice to see statistics that confirm it. Google found that only 56% of the hard drives that failed showed warning signs of failure before they died. So there is a 44% chance you will have no warning at all before your hard drive goes, one day it works, the next day it poofs.

3) Use is not a factor in hard drive failure. Google found no corolation between how much a hard drive had been used, and when it failed. We often hear, “well, but I only use my hard drives on weekends for 15 minutes, so shouldn’t it last forever?” This proves that is not true. Hard drives decay with age, regardless of use.

4) Temperature is probably not a factor in hard drive failure. Google found no corolation between hard drive failure and drive temperature, but the study does acknowledge that the sample set is problematic, because google data centers are all temperature controlled. Previous studies have found a connection between temperature and drive failure, with too cold being more of a factor than too hot.

5) Google found the five year failure rate of hard drives to be 39% with a margin of error of about 9%. This means after five years, somewhere between 30% and 48% of all hard drives have failed. This tracks pretty closely to what we see here in the shop, where 4-6 year old computers often have dead hard drives.

VERY Interesting. I recently returned a 1 TB SATA drive that I plulled
out of a FreeAgent Pro box that "failed" in service as an external
drive for a DirecTV HD DVR. It was slightly under the 5 year warranty.
Seagate is in the process of shipping a replacement. I really didn't
think they would because I had taken it out of the FAP case (I bought
a new 1TB SATA drive and put it in the case), but after reviewing it,
the said they were shipping one out.

In the case of this drive, there was a warning as the performance in
the DVR service kept getting slower. Bacause of that , about a year
ago I started a practice of cloning the entire drive every few months
so I wouldn't lose all the recorded programming. I was fortunate in
that I had just cloned it a few days before the final problems
occurred.

One other point.... even tho I could tell the drive was failing... HD
Sentinel was reporting hundreds of bad blocks.... Sentinel also kept
reoporting the drive was not yet bad enough for a warranty
replacement. After the final problems occurred, I couldn't even access
the drive with HD Sentinel.