1TB sata hard drive?

  • Thread starter Thread starter jw
  • Start date Start date
J

jw

I see a few good buys (?) out there for large internal sata hard
drives. Will a normal BIOS handle 1TB? Will XP SP2? Or are mods
necessary?

Duke
 
I see a few good buys (?) out there for large internal sata hard
drives. Will a normal BIOS handle 1TB? Will XP SP2? Or are mods
necessary?

Duke
Click to expand...

Yes. Yes. No.

Things to watch

1) Not all USB SATA enclosures will work with them. USB
enclosures have strange, unexplained limits.

2) If your motherboard has a SIL3112 SATA controller, there
appears to be a bug in its BIOS code. The motherboard
may hang when it enumerates the 1TB drive. If you plug
in the disk after the system has booted, then the 1TB drive
may work with the OS driver to control the SIL3112.

3) If your motherboard has a VIA Southbridge and SATA 150 port,
use the "force 150" jumper on the back of the drive. Otherwise,
the drive may not be detected, due to a communications problem.
Exception it the VT8237S, which works (my current computer).

4) WinXP has a 2.2TB limit. 2.2TB takes a 32 bit number to
store the sector address. Somewhere in the chain of
events, there is an issue with going larger than that.
There is something called GUID GPT, a different kind of
partition scheme, that can help. Naturally, it might spoil
compatibility with something else. In fact, there is always
some kind of compatibility exposure, once you go over 137GB.
(Like plugging the disk into an older OS that doesn't support
more than 137GB.)

http://en.wikipedia.org/wiki/GUID_Partition_Table

5) 1TB drives may get warm. (There is a fair bit of
variation in the designs.) Make sure the large drive
has good ventilation or forced air cooling.

If you actually put 1TB of data on it, will you be
doing backups ? How long does it take to back up
1TB of data ?

Paul
 
Things to watch

1) Not all USB SATA enclosures will work with them. USB
enclosures have strange, unexplained limits.

2) If your motherboard has a SIL3112 SATA controller, there
appears to be a bug in its BIOS code. The motherboard
may hang when it enumerates the 1TB drive. If you plug
in the disk after the system has booted, then the 1TB drive
may work with the OS driver to control the SIL3112.

3) If your motherboard has a VIA Southbridge and SATA 150 port,
use the "force 150" jumper on the back of the drive. Otherwise,
the drive may not be detected, due to a communications problem.
Exception it the VT8237S, which works (my current computer).

4) WinXP has a 2.2TB limit. 2.2TB takes a 32 bit number to
store the sector address. Somewhere in the chain of
events, there is an issue with going larger than that.
There is something called GUID GPT, a different kind of
partition scheme, that can help. Naturally, it might spoil
compatibility with something else. In fact, there is always
some kind of compatibility exposure, once you go over 137GB.
(Like plugging the disk into an older OS that doesn't support
more than 137GB.)

http://en.wikipedia.org/wiki/GUID_Partition_Table

5) 1TB drives may get warm. (There is a fair bit of
variation in the designs.) Make sure the large drive
has good ventilation or forced air cooling.

If you actually put 1TB of data on it, will you be
doing backups ? How long does it take to back up
1TB of data ?

Paul
Click to expand...

Thanks Paul.

What if I go with a 500GB hard drive instead?
They're pretty cheap now too.

Duke
 
Thanks Paul.

What if I go with a 500GB hard drive instead?
They're pretty cheap now too.

Duke
Click to expand...

If I can get the product information, I prefer to buy
a drive with a single platter. I don't download movies
or music, so my storage requirements aren't that extensive.
My average disk here is 80GB, and the last two I bought
a couple months ago, were 250GB. Buy whatever suits your
projected storage requirements. Use the reviews on Newegg,
for the various product possibilities, to spot any high
failure rate items.

Paul
 
(e-mail address removed) wrote:

What if I go with a 500GB hard drive instead?
They're pretty cheap now too.

Duke
Click to expand...

It should depend on how much disk space you need. My current system has
3TB and I think I can enjoy few more.
 
Joel said:
(e-mail address removed) wrote:



It should depend on how much disk space you need. My current system has
3TB and I think I can enjoy few more.
Click to expand...

How do you back up those 3TB ? How long does
it take for the backup to run ?

Paul
 
Paul said:
Yes. Yes. No.

Things to watch

1) Not all USB SATA enclosures will work with them. USB
enclosures have strange, unexplained limits.

2) If your motherboard has a SIL3112 SATA controller, there
appears to be a bug in its BIOS code. The motherboard
may hang when it enumerates the 1TB drive. If you plug
in the disk after the system has booted, then the 1TB drive
may work with the OS driver to control the SIL3112.

3) If your motherboard has a VIA Southbridge and SATA 150 port,
use the "force 150" jumper on the back of the drive. Otherwise,
the drive may not be detected, due to a communications problem.
Exception it the VT8237S, which works (my current computer).

4) WinXP has a 2.2TB limit. 2.2TB takes a 32 bit number to
store the sector address. Somewhere in the chain of
events, there is an issue with going larger than that.
There is something called GUID GPT, a different kind of
partition scheme, that can help. Naturally, it might spoil
compatibility with something else. In fact, there is always
some kind of compatibility exposure, once you go over 137GB.
(Like plugging the disk into an older OS that doesn't support
more than 137GB.)

http://en.wikipedia.org/wiki/GUID_Partition_Table

5) 1TB drives may get warm. (There is a fair bit of
variation in the designs.) Make sure the large drive
has good ventilation or forced air cooling.

If you actually put 1TB of data on it, will you be
doing backups ? How long does it take to back up
1TB of data ?

Paul
Click to expand...
Allow me to complicate the issue somewhat.
This will be my first SATA device, so be gentle with my ignorance.
Thought I'd buy a 1.5TB SATA drive just for fun.

I use plug-in PATA drives with operating systems from dos to win7-beta
and several flavors of linux.
Back in the day, I could plug drives in and out with abandon.
I had to give up on the second drive when XP arrived.
Problem was that somewhere in the middle of the boot process,
XP decides that it can't find the boot drive. I'm guessing that
the presence of the second drive messed up some map stored in the
registry. So, XP pulls the rug out from under itself.

I never could make it happen on command, but sometimes, XP would
corrupt itself and refuse to boot ever again, to protect me...sorry
for any inconvenience.

My motherboard has SATA ports:
South Bridge 82801EB/ER (ICH5/ICH5R) LPC Interface Bridge

Since big drives only come in SATA, I'll want to put it inside the box.
Don't want to boot from it, just use it for bulk storage.
Yes, I understand that older operating systems won't see it, but
I can live with that...as long is it's not corrupted.
I'm unwilling to go into the bios and turn off the sata controller
every time I switch PATA drives. OS needs to just ignore the drives
it doesn't have drivers for.

My Plug-in PATA drives have up to four primary partitions.
If I were to add a second PATA drive, drive letters get moved
around. At best it breaks software. At worst, sometimes,
it breaks the OS, permanently.

Question is this...
When I add the SATA device, is it gonna move drive pointers around
(and windows drive letters)
and break all my M$ OS installations?

USB-connected drives don't have this problem, but I'd like
to reduce the clutter on my desk, not add to it.

Suggestions?
Thanks, mike
 
spamme0 said:
Allow me to complicate the issue somewhat.
This will be my first SATA device, so be gentle with my ignorance.
Thought I'd buy a 1.5TB SATA drive just for fun.

I use plug-in PATA drives with operating systems from dos to win7-beta
and several flavors of linux.
Back in the day, I could plug drives in and out with abandon.
I had to give up on the second drive when XP arrived.
Problem was that somewhere in the middle of the boot process,
XP decides that it can't find the boot drive. I'm guessing that
the presence of the second drive messed up some map stored in the
registry. So, XP pulls the rug out from under itself.

I never could make it happen on command, but sometimes, XP would
corrupt itself and refuse to boot ever again, to protect me...sorry
for any inconvenience.

My motherboard has SATA ports:
South Bridge 82801EB/ER (ICH5/ICH5R) LPC Interface Bridge

Since big drives only come in SATA, I'll want to put it inside the box.
Don't want to boot from it, just use it for bulk storage.
Yes, I understand that older operating systems won't see it, but
I can live with that...as long is it's not corrupted.
I'm unwilling to go into the bios and turn off the sata controller
every time I switch PATA drives. OS needs to just ignore the drives
it doesn't have drivers for.

My Plug-in PATA drives have up to four primary partitions.
If I were to add a second PATA drive, drive letters get moved
around. At best it breaks software. At worst, sometimes,
it breaks the OS, permanently.

Question is this...
When I add the SATA device, is it gonna move drive pointers around
(and windows drive letters)
and break all my M$ OS installations?

USB-connected drives don't have this problem, but I'd like
to reduce the clutter on my desk, not add to it.

Suggestions?
Thanks, mike
Click to expand...

It could be, that when you installed WinXP, you had the other drive
present as well. I disconnect the other drives before doing
an install. If I needed to manage all the OSes from one place, I
might use Boot Magic from my Partition Magic CD. It allows a
user to add entries to the menu, after an OS is installed on
another disk. There are many other boot managers, but they tend
to be "greedy", and update their menu for you, and somethings
that can go horribly wrong. I like to fix up the boot issue
myself, after the dust has settled. My current machine isn't
managed by Boot Magic, and I use the F11 key to select a
boot drive from the BIOS. So my BIOS is my boot manager right
now.

In terms of BIOS issues, on my current computer, the BIOS
drives me crazy. The BIOS has its own scheme of numbering
drives, and if you install Linux, the Linux installer uses
the BIOS numbering scheme. This caused me to have the MBR
on my WinXP disk erased, when installing Linux on a completely
different disk. As a result, I recommend disconnecting
everything that might get trashed by a well written installer :-(
I don't really understand what is going on, and why the
BIOS labeling of drives, has to change as a function of
which drive is the nominal boot drive.

When you add a drive, it is going to need a drive letter for
every partition that you prep and format. The letters can be
assigned manually, using Disk Management. So you can lift them
up high if you want. For example, I have a disk at S:, whose
label is Scratch. I have a certain number of partitions that
fall where they may, down low. And as a matter of practice, right
after I install an OS, I always move the CDROM out of the way as
well. Perhaps move the CD to Q:. Since I do that immediately,
if I then install Microsoft Office right afterwards, the registry
will be filled with references to Q: instead of some other low
drive letter like D:. Your CD should be in a stable place, so that
there won't be an issue with references to that optical drive later.
By picking Q:, my assumption is that the drive won't get bumped
due to other randomly placed volumes.

In terms of compatibility, if you have some OS which does not
understand 48 bit LBA (>137GB), then that OS could corrupt
the disk if it manages to recognize a partition (NTFS or FAT32
say) and attempts a write to them. So that is an exposure you should
consider. I have to be careful here, as I have one out-of-date
OS frozen at a certain Service Pack, and if that OS writes
to my 250GB SATA drive, there could be trouble. Right now,
all the partitions on that drive are below 137GB. And that
makes it safe to boot that OS, with the SATA drive connected.

I've read that USB devices can have drive letter problems.
For example, network shares and USB drive letters don't know
about one another. A network share can grab the drive letter
from a USB device. Uwe Sieber's site has more details. He also
has some good info on some issues with USB and Windows.

http://www.uwe-sieber.de/usbdlm_e.html
http://www.uwe-sieber.de/usbstick_e.html

Paul
 
Paul said:
It could be, that when you installed WinXP, you had the other drive
present as well. I disconnect the other drives before doing
an install. If I needed to manage all the OSes from one place, I
might use Boot Magic from my Partition Magic CD. It allows a
user to add entries to the menu, after an OS is installed on
another disk. There are many other boot managers, but they tend
to be "greedy", and update their menu for you, and somethings
that can go horribly wrong. I like to fix up the boot issue
myself, after the dust has settled. My current machine isn't
managed by Boot Magic, and I use the F11 key to select a
boot drive from the BIOS. So my BIOS is my boot manager right
now.

In terms of BIOS issues, on my current computer, the BIOS
drives me crazy. The BIOS has its own scheme of numbering
drives, and if you install Linux, the Linux installer uses
the BIOS numbering scheme. This caused me to have the MBR
on my WinXP disk erased, when installing Linux on a completely
different disk. As a result, I recommend disconnecting
everything that might get trashed by a well written installer :-(
I don't really understand what is going on, and why the
BIOS labeling of drives, has to change as a function of
which drive is the nominal boot drive.

When you add a drive, it is going to need a drive letter for
every partition that you prep and format. The letters can be
assigned manually, using Disk Management. So you can lift them
up high if you want. For example, I have a disk at S:, whose
label is Scratch. I have a certain number of partitions that
fall where they may, down low. And as a matter of practice, right
after I install an OS, I always move the CDROM out of the way as
well. Perhaps move the CD to Q:. Since I do that immediately,
if I then install Microsoft Office right afterwards, the registry
will be filled with references to Q: instead of some other low
drive letter like D:. Your CD should be in a stable place, so that
there won't be an issue with references to that optical drive later.
By picking Q:, my assumption is that the drive won't get bumped
due to other randomly placed volumes.

In terms of compatibility, if you have some OS which does not
understand 48 bit LBA (>137GB), then that OS could corrupt
the disk if it manages to recognize a partition (NTFS or FAT32
say) and attempts a write to them. So that is an exposure you should
consider. I have to be careful here, as I have one out-of-date
OS frozen at a certain Service Pack, and if that OS writes
to my 250GB SATA drive, there could be trouble. Right now,
all the partitions on that drive are below 137GB. And that
makes it safe to boot that OS, with the SATA drive connected.

I've read that USB devices can have drive letter problems.
For example, network shares and USB drive letters don't know
about one another. A network share can grab the drive letter
from a USB device. Uwe Sieber's site has more details. He also
has some good info on some issues with USB and Windows.

http://www.uwe-sieber.de/usbdlm_e.html
http://www.uwe-sieber.de/usbstick_e.html

Paul
Click to expand...
Thanks for the links.
I already do most of what you suggested.
I don't have any problem working with the system
after it boots. I do have a problem if it won't boot.
And I have a REAL problem if it trashes the disk.

Lemme see if I can shed some more light on the subject.
I'd like to hear how it works now, but in the old days,
the bios and/or OS assigned drive letters thusly.

Assume drive 1 and drive 2 each have partitions A and B.
If you plugged in drive one you got
C: = 1A D: = 1B

If you plugged in two drives, you got
c: = 1A D: = 2A E: = 1B F: = 2B

So, if your boot manager reassigns drive letters,
adding a second drive messes all that up.
If you're expecting to boot from 1B, you won't get what
you expect.

XP adds the further complication by remembering how
to assign new drive letters. I don't know if this
info is stored on the drive or in the registry.
But if you move the physical addresses of the drives,
that info is no longer valid.
Sometimes, it just locks up on boot.
Sometimes I get a stop errror.
Sometimes the system gets trashed to the point
that I can never get XP to boot again, no matter how
I rearrange the drives.
I got so tired or reinstalling the os that I quit
swapping in a second drive or using a boot manager.
And I quit experimenting to determine how to fix it.

Back to the present.
If I connect the SATA drive (and always leave it conntected)
and reinstall all the operating systems, I'd expect it to work fine.

Problem that I'm worried about is adding the SATA to the
existing mix and having stuff break.
 
How do you back up those 3TB ?
Click to expand...

Having just recently added my 3rd TB to my primary storage server's
largest array, my answer is that I don't backup, I replicate.

I have two redundant file servers which mirror all data, offering me
redundancy against drive failure, controller failure, or even
OS/motherboard/power-supply failure on either server.

Both servers also create shadow copies which allow me to return to a
point in time in the past, protecting me against unwanted deletions or
modifications.

Depending on the value of the data and expected frequency of change,
we're creating shadow copies anywhere from 3 times a day (lunch, end of
business, 3am) to every 3-4 days (for less critical or easily
replaceable data) -- I usually configure the two servers to perform
shadow copies at different times, increasing the odds that a file
created and subsequently lost today can be recovered.
How long does
it take for the backup to run ?
Click to expand...

With my setup, the servers are linked using gigabit ethernet, so the
answer is somewhere around the amount of time it takes the data to
transfer at 400Mb/s.

One of the major benefits of this system is that in the event of a
failure I don't lose data between the most recent backup and now, that
data is on the redundant server almost immediately.
 
Back
Top