Booting Win7PE from RAM

I've a few recovery environments which are Win7PE based. Each of them come with their own BCD datastore.I copied the BOOT,SOURCES folder & BOOTMGR to an UFD & placed them in the root. I chainloaded BOOTMGR from Grub4DOS. The error displayed was -

Status: c00000f

Info: The boot selection failed because a required device is inaccessible.

Below is the contents extracted from the BCD of Norton Ghost 15 rescue CD

Windows Boot Manager

--------------------

identifier              {bootmgr}

description          Windows Boot Manager

locale                  en-US

inherit                 {globalsettings}

default                 {default}

displayorder            {default}

toolsdisplayorder       {memdiag}

timeout                 30



Windows Boot Loader

-------------------

identifier              {default}

device                  ramdisk=[boot]\sources\boot.wim,{7619dcc8-fafe-11d9-b411-000476eba25f}

path                     \windows\system32\boot\winload.exe

description         Windows Setup

locale                  en-US

inherit                 {bootloadersettings}

osdevice             ramdisk=[boot]\sources\boot.wim,{7619dcc8-fafe-11d9-b411-000476eba25f}

systemroot         \windows

detecthal             Yes

winpe                  Yes

ems                     Yes

What adjustments do I need to do?

I've a few recovery environments which are Win7PE based. Each of them come with their own BCD datastore.I copied the BOOT,SOURCES folder & BOOTMGR to an UFD & placed them in the root. I chainloaded BOOTMGR from Grub4DOS. The error displayed was -
[code=auto:0]Status: c00000fInfo: The boot selection failed because a required device is inaccessible.

I have also been playing with things like this - i.e. more than one "wim" file that I wanted to have available...There are a few other posts that discussed this (just do some searching)...They led me to conclude that you can either edit the BCD file and make the menus available in the bootmgr - OR take the approach of this tutorial over on the RMPrep USB site.I took the latter approach for a script to add in the ActiveBoot disk...**BUT** I would still see this same sort of error when the PE I was trying to create (i.e. packing the wim file and editing the main BCD) after I had put the additional (and modified) boot1.wim and BC1 in place...I just gave up trying to figure out what part of the process of packing the wim, etc. (i.e. the things in the "finalize stage") was causing this to occur, but since It worked OK if I added the files AFTER the boot.wim for the win7PE_SE was created, I just made my script execute after that...Maybe something like that could be useful for you...PM me if you want a copy of my script...(still in development)

I also know that there is a script that automates some of the BCD editing called "addmorewim.7z" that is on this forum somewhere...so do a search for it...Maybe that can also give you sme clues on what to edit...


Good luck
Scott

Some links that I've found out are
http://technet.micro...845(WS.10).aspx

http://keznews.com/2...stored_WIM_file

http://technet.micro...195(WS.10).aspx

What adjustments do I need to do?

What does this string mean?

{7619dcc8-fafe-11d9-b411-000476eba25f}

Try reading attentively this:
http://www.multiboot....uk/floppy.html

Particularly:

The BCD file on the Vista floppy has to be specific to the hard drive that Vista is installed on. You can make the rest of the floppy on another computer, but you need to use the BCD file from the machine the floppy is for, (unless you manually customise one by re-specializing the BCD boot Objects). If you have 'generalized' the whole BCD in the manner described on the Cloning page then it will not work.

Then, read here:
http://www.multiboot...ml#respecialize


Wonko

I'm stuck @ another point. The original problem I described in the first post is resolved. I simply repeated the process I did & it's NOT there anymore. It occured, I believe, the snapshot of the BCD I posted & the one I used - were diffreent. It was my mistake, I apologize.

The above tests were done on VMware with a VM equipped with 192MB of RAM. So, now it gives a misleading error - "NDIS.SYS is corrupt". I've come acroos this error earlier also. But, I've no option to increase the VM's RAM size(@ Wonko : No, the new yr has no impact on the VM, don't know about myself yet

@Wonko

What does this string mean?

The strange string is basically an effort to convey the query to the rebooters asking that if I require to do some offline adjustments to the BCD store, what should they be?

--------------------------------------------------------------------------------------------------------------------------------

Now comes the most critical/ astonishing part. I've already posted here what problem I faced in transferring WIn7setup to UFD using WinSetupFromUSB 0.2.3 BETA. Nobody could give a satisfactory reply. Initially I thought it to be a problem pertaining to WinSetupFromUSB 0.2.3 BETA. To be sure, I tried out the following:

• Formatted the 2GB Transcend UFD with RMPrepUSB v 2.0.811 selecting the FAT32 radio buttoon & checking the Boot as HDD checkbox.
• Installed Grub4DOS by hitting the Install Grub4DOS button.
• Copied BOOTMGR to the root of the UFD.
• Created a folder called BOOT & copied the BOOT.SDI & BCD from the Norton Ghost 15 recovery ISO.
• Created a SOURCES folder & copied the BOOT.WIM into it from the same source.
• Chainloaded BOOTMGR from Grub4DOS.

The UFD prepared above booted fine in VMWare, @ least the initial part. But when I tried out the same on REAL HARDWARE (my desktop), it produced the same error message as pointed in the mentioned thread.

Windows Boot Manager



A recent hardware or software changes ...............



<Some general instructions>



File: \BOOT\BCD

Status: 0xc000000e

Info: An error occurred while attempting to read boot configuration data



ENTER=CONTINUE                     		ESC=CANCEL

Then I deleted the /BOOT/BCD - but the error message was the same. It essetially means BOOTMGR is unable to read BCD store from the disk. Am I the only facing this problem? I mentioned it to let people know that actually we are fighting against a coule of problems here as opposed to one I mentioned earlier.

It essetially means BOOTMGR is unable to read BCD store from the disk. Am I the only facing this problem?

That's a common error at some BIOS. I don't know the true reason.
Work around: boot a floppy, disk or ISO image
Compare http://www.911cd.net...ndpost&p=137270

The strange string is basically an effort to convey the query to the rebooters asking that if I require to do some offline adjustments to the BCD store, what should they be?

You were given (meaningful ) pointers.
What about READing them and checking them against your problem?


Wonko

What about READing them and checking them against your problem?

I'm in the process. Surely report u back.

That's a common error at some BIOS. I don't know the true reason.
Work around: boot a floppy, disk or ISO image
Compare http://www.911cd.net...ndpost&p=137270

Thanks a lot cdob. A nice solution was there. I just introduced a small variation. Instead of

Use WinImage to set boot sector of vistape.ima (boot sector properties) to Vista Bootmgr.

I directly chainloded the BOOTMGR as it seems to be an additional/ unnecesary step to manipulate the bootsector. Everything went fine.

If u or somebody can find out the true reason of the problem, please let me know.

I directly chainloded the BOOTMGR as it seems to be an additional/ unnecesary step to manipulate the bootsector.

I prefer a ISO image nowadays.
Or course you may use another image supported by grub4dos.

If u or somebody can find out the true reason of the problem, please let me know.

As for USB boot and error message:

File: \Boot\BCD
Status: 0xc000000e
Info: An error occurred while attempting to read the boot configuration data

Example HP DX2000MT, grub4dos not used: http://www.msfn.org/...post__p__847479

So let's talk about how this specific motherboard works. If at any point, the firmware re-enumerates any of the devices, this could definately happen. For example, let's say that at a certain point of the boot process it re-enumerates the bus that the USB device resides on. This would reset all devices on that bus. If the computer was in process of booting off of something connected to that bus, those devices (as far as the booting OS would be concerned) would have disappeared. This is the case in those times if PnP is not enabled (such as the early boot process of the PE) and while the device is (usually instantly) ready on the system, the PE will not see it until PnP is enabled. Since it cannot see the device, it can't finish loading. This is my guess about the situation since I've also experienced a couple odd-ball boards that can't boot.

I've a D845GEV2:
\Boot\BCD is missing at USB boot.
Booting PE1 flat files result to a BSOD 0x7B. Does boot with Dietmar's ntdetect.com.

Can you verify:
Boot a PE1 flat files from USB at your machine, use a default XP ntdetect.com.
Do you get a BSOD 0x7B?
Add Dietmar's ntdetect.com. Can you boot PE1 now?
http://www.911cd.net...indpost&p=90483
http://www.msfn.org/...ndpost&p=738009

I understand: ntdetect.com feature is included in bootmgr nowadays.
Can anybody adjust bootmgr?

I prefer a ISO image nowadays.
Or course you may use another image supported by grub4dos.

Why ISO image is better than floppy image?

Can you verify:
Boot a PE1 flat files from USB at your machine, use a default XP ntdetect.com.
Do you get a BSOD 0x7B?
Add Dietmar's ntdetect.com. Can you boot PE1 now?

You are ABSOLUTELY right. Default NTDETECT.COM produces a BSOD while Dietmar's one boots normally. Correct guess cdob.

Correct guess cdob.

I would call it excellent detective work!


Wonko

I would call it excellent detective work!

One of the main reasons for me of enjoing my stay here is to have the grump, witty, old, bearded (don't know if he's still so), boo-land Italian ex-Admin as my neighbour.

@Holmes.Sherlock
From a PM from joakim (whom I had made aware of the current issue asking him to have a look at the matter if possible ):

I was lucky enough to borrow a computer with this horrible AMI BIOS. This is version 07.00T from 2002 and it produces the same error message (0xc000000e) as you and many others reported. But on this particular computer I've found a working solution. Use RMPrepUSB with option 1 (bootmgr) and FAT16 formatted (ie no chainloading). Verified on 2 sticks at 256 mb and 2 gb.

Trying to chainload the same working sticks from grldr on local HDD, produces same error as above. Sticks formatted with FAT32 or NTFS will also not boot. NTFS produces "Disk read error", and FAT32 will just hang with cursor blinking. Chainloading those sticks via grldr also does not work, and chainloaded NTFS stick will also throw 0xc000000e, whereas chainloaded FAT32 will give "unmounable something... volume blabla".

So I still cannot reproduce the error, as I found a solution on this particular machine.
However, maybe it's a solution to the issue with AMI bios, usb and bootmgr...

Could you give it a try?

If you still have that PC around, please try the FAT16+standard MBR/bootsector and report.


Wonko

If you still have that PC around, please try the FAT16+standard MBR/bootsector and report.

The PC is my own desktop(with an INTEL D915GAV MOBO), so I'll give a try & report back.

BTW, as far as USB booting is concerned, I guess, the BIOS vendors can't come to a consensus & that's why there is nothing what we call "standard". But, when the creation of inconsistency comes, do they set up a "round-table-conference" to discuss amongst themselves how to make customers' life a hell?

The PC is my own laptop, so I';; give a try & report back.

BTW, as far as USB booting is concerned, I guess, the BIOS vendors can't come to a consensus & that's why there is nothing what we call "standard". But, when the creation of inconsistency comes, do they set up a "round-table-conference" to discuss amongst themselves how to make customers' life a hell?

Naaah.
They are good guys after all, they don't do it (I mean sucking code in BIOS) on purpose, they do it because of:

• ignorance
• stupidity
• stupid time deadlines (or budget ones) imposed by their stupid management
• continuous turnover so they don't know or don't understand enough or have not the time to study and understand what the codebase their predecessor made and left them did, why, and how and simply re-use that code
• sloth

... and there are far worse BIOSes than yours ....

Anyway I'm pretty sure that nothing (not even good BIOSes) are made "on purpose" or "by design", that is what is used to either take the merits of a good product or to lessen the blame for a bad one.

BIOS code (and conversely a lot of other things in modern world) are governed by chaos or randomness much more than they are by "logic" or what I call "senseless evolution", an approach where the final customer has the least importance.

If you have a product that sucks instead of bettering the product, you hide it's drawbacks and pump up it's "enhanced status" (over a previous perfectly working product) and additional (often senseless) "added features".

I still would like to meet the guys that decided that a TV set needed an OS to run on , BTW.
http://reboot.pro/9915/

And yes, next post will be by someone that will accuse me of being against new ideas, I know.


Wonko

And yes, next post will be by someone that will accuse me of being against new ideas, I know.

No, u r wrong. The next post is mine & not going to accuse u anyway.

Now, time 2 report the result. This is what I did, step-by-step :-

• Formatted my transcend 2GB Jetflash UFD with RMPrepUSB v 2.0.811 selecting WinPEv2/WinPEv3/Vista/Win7 bootable [BOOTMGR] (CC4) radio button in Category - 3 (BOOT OPTIONS) & selecting FAT 16 radio button in Category - 4 (FILESYSTEM & OVERRIDES). I didn't check any check box in the second category.
• Copied BOOTMGR from WIN 7 installation image to the root of the drive.
• Copied BOOT folder from WIN 7 installation image to the root of the drive.
• Created a folder called SOURCES @ the root of the drive.
• Copied BOOT.WIM from WIN 7 installation image to the /SOURCES folder.

The resulting stick booted fine in VMWare but, when tested on my desktop, it produced the same 0xc000000e error.

I presume you also have to check "boot as HDD (C: 2 ptns)"

Can you confirm that we are talking of the same AMI BIOS joakim tested successfully with both 256 Mb and 2 Gb sticks?

It is well possible that this workaround only works on that.

... at least we tried...


Wonko

I get,
2 GB USB stick (the smallest I pocess)
D845 PEVS, RMPrepUSB and FAT16.

Still 0xc000000e missing /boot/bcd file.

True reason is unkown still.
CHS / LBA mapping or another USB reason.

@Steve
Do you possess a relating machine?

Can you confirm that we are talking of the same AMI BIOS joakim tested successfully with both 256 Mb and 2 Gb sticks?

The awkward point about my desktop's BIOS is it doesn't show up its manufacturer's name in any of its pages while my laptop's BIOS does.

The awkward point about my desktop's BIOS is it doesn't show up its manufacturer's name in any of its pages while my laptop's BIOS does.

Well , you DO NOT normally look for the BIOS maker/version on "any of it's pages".
You (deftly and quickly ) press the [Pause] key when the pc is booting (AFTER having turned off "graphical boot" - if any - in the BIOS setup).
Read here:
http://www.wimsbios.com/
http://www.wimsbios.com/faq.jsp
http://www.wimsbios....anufacturer.jsp


Wonko