Vista: Are there issues with the disk ID?

Cheers guys, it was a bit of a project that I never really finished to my satisfaction. Life got in the way. There were a couple of other pages planned – one of which was going to be on the very subject of the Vista created partition offset. There is also a couple of other issues I've yet to see mentioned anywhere (they have nothing to do with the topic of this thread – I don’t think) and plenty of other points to be expanded on, as well as a bcdedit.exe guide for creating new GUIDs. Hopefully I’ll be able to get back to it shortly.

 

I have a coupla' comments about "non-technical users" as they've been discussed on this thread. So, my post here is NOT a post on 'Vista: Are there issues with the disk ID".

Mustangs discussion on page 2 of this thread alluded to these "non-technical" users, and Mustang expressed an admirable sympathy for these folks and castigated the software manufacturers marketing for leading them down the rosy path to disaster - especially if they chose to do a restore without technical help, which is the way all this disk imaging software is marketed. The marketers represent their software as something that even the most inexperienced user can use with ease - which IS NOT the case. So, I definitely agree with Mustang's analysis there.

But I'm afraid that I don't have the same patience and good will for non-technical users. Here's why:

First of all, I'm certainly not an "advanced user" myself, maybe a "beginner intermediate", a little bit of a geek, but in the company of the technical minds of some of the users here, like Mustang and a few others on this thread, I'm really an amateur.

Nevertheless, I do give lessons to some of the Senior Citizens (I'm one too) in our community on how to get around in an OS and some other basic computing skills (VERY basic). Now at this point, you may be asking "Geezzz, if you don't have much patience with non-technical users, what the heck are you doing trying to teach them technical stuff??". Good question, and there's a simple answer - I HOLD MY TONGUE!!!! But under my breath I'm thinking and saying "Hey, Lady/Guy, your dummer'n dirt, so you shouldn't be driving one of these things unless you have a license!!". So far, I haven't said that out loud, and when I do I'm sure the Community Center will ask me to stop being a volunteer.





I also troubleshoot some computer systems for my Senior Citizen neighbors. In that effort, I see systrays just filled with all manner of little icons, most of which will automagically start up the software when these folks turn on their systems, and then keep it running as a TSR. And then they ask me why their systems take so long to boot (well . . . they usually don't use the word "boot"). Sheeesh!!! When I point out to them that they have all this crap loading and point out that it's crap they don't need or rarely use, and when they use it can just as easily run it from the Start menu, they look at me like I just spoke to them in Greek.

<digression> Now I said that compared to you guys on this thread, I'm not really all that technical, but when I go to my users, they think I'm a genius. In the land of the blind, the one eyed man is King. <end digression>​

And some of them (most of them, actually) don't even know what software most of those systray icons run.
Which brings me to my next point, and one which Mustang and Tabvla touched on - marketing and marketing ENGINEERS (there's an oxymoron if there ever was one). Today's software comes out of the box with the "Start at start-up" option enabled - which is why most people have their systrays full of icons. When they install the software, most using the "Complete" install option instead of the "Custom" install option, they install that "Start at start-up" option and then get all those tiny icons in their systray.

Granted, a few icons in your systray are appropriate - like I always see my blue, white and red shield for McAfee 8.0i in my systray. But when your systray is overflowing with them

Now once I tried to explain this "Start at start-up" option, and why you don't need this most of the time, but my user interrupted with "Aren't computers supposed to do stuff automatically so that you don't have to do it??". I started into my explanation about using the Start Menu, and about how "automatic" can be a bad thing sometimes (like some of those M$ "automatic" updates when they would corrupt other stuff), but I saw their eyes glazing over and the deer-in-the-headlights stare, and I stopped.

Let me try to distill this at this point. Today's software is marketed not only that it will do what it says it will do for users with no technical background, but it will also do it . . . AUTOMAGICALLY!!! And that "automatic" thing is really one of the primary reasons most software these days screws up, and why non-technical users get so frustrated when their systems malfunction. Because that "automatic" stuff is what makes most software incompatible with other software. When things go "automatically", they don't necessarily check for incompatibilities with other installed software. That's why users SHOULD be technical, so that they can check for software incompatibilities. And another reason that most software these days screws up is because those marketing guys want to rush their software to the shelves BEFORE it's ready for prime time (a point I think Tabvla made).

It's a Catch-22. Software makers are putting out stuff that does more things automatically because that's what users buy and . . . THINK . . . they need, when it's precisely that very "automatic" behavior that screws up their systems!!!!

The first thing I do when I install new software is disable the "Start at start-up" option and disable all the automagic stuff. For example, I do my (daily now because of all the malware out there) DAT updates for my antivirus software manually, and I try to teach people how to do those updates manually. I can't tell you how many people think that just because they think antivirus updates are automatic, that their antivirus software is up to date, or just because they see their antivirus software icon in the systray they think everything is OK. Not too long ago, I found one guy with McAfee 4.0 and he thought that just because he saw the shield in his systray he was OK - that 4.0 hasn't been around since about a decade ago, and the scan engine isn't even compatible with DATS from several years ago.

OK . . . end of rant on "non-technical" users.

Oh . . . one more thing. When I try to teach these folks how to navigate with a mouse, I'm always saying "Get your hands off the mouse when you're not ready to click" because those random clicks from a hand resting on the mouse and twitching are what gets people into trouble. (I think someone on this thread mentioned that).

 

This is kind of funny in a way.

I installed Vista Premium on my PC on 4/18/07. The first software I installed on it was TI 10 latest build for Vista, so that I could get an image of my fresh installation.

All went well, and over the next couple of days I continued installing software until I had just about everything loaded, making a couple of backup images along the way, from within Vista, to an external hard drive.

On the third or fourth day I installed and ran "Vista Compatible" jv16 Powertools. Completely hosed my registry.

Now here's the funny part, I had not read the forums regarding potential TI/Vista problems, and completely oblivious to any potential disaster, and having complete confidence in TI, I simply booted to my TI boot disk, pointed it to my most recent backup, chose restore with mbr, and voila! I'm using that restore right now!

While I consider myself to be capable with computers, I am not nearly as savvy as most of you folks. In fact I'm still having problems making TI do my bidding as regards limits on number of backups allowed in "Backup Places".

So there's my two cents worth for whatever it's worth.

About two cents I imagine!

Dell E510
Pentium D 820
2GB DDR2 SDRAM
Dual 250GB Seagate SATA HDD
Windows Vista Ultimate

 

Just as a point of interest, I can tell you that both Norton Ghost 12 and Symantec Backup Exec 7 Desktop Edition are able to handle the Vista 2048 sector partition offset. No repair is needed and the 2048 sector offset is maintained after the restore.

 

Hmmmmm .... interesting.

 

I mailed acronis and they told me that build 3887 of Workstation version should fix this.. wel.... NOT!!!

Im working an a Vista Boot Fix plugin For Bartpe. Will post it later.

This plugin wil only work for Single Bootdrives where Vista is on C:
Maybe that people with more script skills kan enhance the plugin to also work with multi boot systems........

stay tuned!

 

Here it is:

Tested and WORKING!

https://www.wilderssecurity.com/showthread.php?p=1005755#post1005755

 

Hi Bornconfuzd,

Thanks for your feedback..... which is certainly worth more than two cents!

In your post you wrote: -

In my experience this is normal, the reason being that you restored the image to the same disk and the same partition on which Vista was originally installed. I have not personally experienced a problem with doing this, although other members may have.

My original concern however still remains. Vista uses the Disk ID as an important parameter in the boot process. When Vista is installed it incorporates the Disk ID of the disk on which it is being installed into the boot configuration data. If an image of that installation is restored to a different disk Vista will abort on the first boot and will require a repair.

This is a serious issue because, if the system disk fails, and you need to restore your image to a new disk, you may be left with an unusable system if you do not have the necessary technical skills or software tools to perform a repair.

The entire reason for the existance of software such as Ghost or True Image is to enable an ordinary user to recover their system in the event of a catastrophic failure - which must include the possiblity of restoring to a new disk.

T.

 

Hi Tabvla,

Thanks for your kind remarks.

I've gone back and reread ALL the posts though, and have a better understanding of the concern.

At first I thought it was for both, making a total restore to the original HDD, and restoring to a new disk.

If I'm correct though, it seems that the main concerns are for restoring to a new disk, and restoring a single partition.

I've spent many hours reading the posts and links on this forum, and have learned a great deal. In fact, I stumbled onto this thread while looking for an answer to another question I had.

Also I must say, I'm quite grateful to all you folks for sharing your technical expertise. People like myself would never be able to properly use our software, if at all, without your help!

For others like me, who might read this post, part of my backup strategy, since the dos days, is to use plain old "Copy". I simply copy my personal data files to an external HDD on a regular basis, about once a week. Then in the worst case scenario, if I must do a clean installation, I can just copy my data back. Primitive, but effective.

I'll be keeping a close watch on this thread though, because I have been thinking about installing a new, larger hard drive and using TI 10 to image it.

Thank you all again for all you do.

 

Ran a few tests MudCrab and can see what’s going on, but why it still not quite clear. It is only the first primary partition that has its offset moved due to the stated “future large sector hard drive support”. It’s another factor that is affecting following primaries. All Vista created partitions are ignoring the convention of aligning with cylinder boundaries and instead using another criterion. This puts the ending sector of any partition short of the nearest cyl boundary (in all my tests so far by 4888sectors). A Vista partition following a conventional primary will have its offset a bit past the cyl boundary. Acronis and most other partitioning tools – including those in XP – always default to the nearest cyl boundary for both the start and end of a partition, hence the problem. True Image is obviously aligning its restored Vista images to cyl boundaries.

The first logical in an Extended partition also has the starting offset of 2048 sectors, following logicals are aligned by the new criteria.

Finding the exact reason for the new partition aligning has proved allusive. The best clue I think comes form this chat session http://blogs.technet.com/filecab/pages/448274.aspx

I can’t figure out why it needs 4888 sectors to do this, so perhaps there is more to the story. This gives no real clue http://forums.microsoft.com/TechNet/ShowPost.aspx?PostID=994065&SiteID=17



This is unrelated – but worth a mention http://www.dcr.net/~w-clayton/Vista/DisappearingPartitions/DisappearingPartitions.htm I can confirm these results, but I doubt it has to do with the partition itself or the new aligning rules.

 

McTavish,

Thanks for the links. It is interesting to see what's going on.

Also, note that it is not only the first primary partition that gets a larger offset. As I noted in post #34, the second primary partition also had a larger offset (probably to align as noted in your post), though it was not not as large as the first partition.

I have not run any tests to confirm if access is actually faster with the default Vista alignment compared to the "old" default. My guess is that this is actually minimal, but I could be wrong. If it did end up being a 50% performance improvement, then just restoring a TI image and doing the repair to fix the offset problem will lose it. And that wouldn't be good.

Personally, it seems silly to me to force this new "offset" onto today's drives just because tomorrow's larger drives will need it. When a user needs to upgrade, I doubt it will make any difference whether the old "smaller" hard drive had the offset or not. The new drive will be formatted as required and the restore/install will continue from there. In any case, all third-party partitioning/imaging programs will have to be updated to support it on all drives.

Another thought, I wonder if Acronis is going to support restoring a Vista image that had the "old" offset to a partition with the "new" offset to preserve the Vista standard and possible performance benefits. In other words, if the partition had the 63 offset and was imaged, then restored to the same drive but changed so the offset was 2048.

 

Any Vista created partition will have an offset that is not on a cyl boundary, whether that offset is larger or smaller will depend on the preceding partition. If the partition in front of it was also created by Vista, then the offset will be smaller (by 4888 sectors) because the new partition will start directly after the first partition, with no empty sectors at all between partitions. If the preceding partition ends on a cyl boundary then the Vista partition offset will be larger and there will be unused sectors between partitions. What the default for the “Cache Aligned” larger offset is I did not look at yesterday. If I get time today I will run tests on various sized hard drives and different machines to see if there is a standard default value.

If anyone here knows anything about this “Cache Aligned” thing or has the time to research it then I would be grateful for the knowledge. Surely it can’t just be the same as cluster aligning partitions to improve I/O performance as the default cluster is only 8 sectors. http://geekswithblogs.net/ntpro/archive/2005/08/11/49948.aspx
http://www.msexchange.org/tutorials/Disk-Geometry.html

 

Guys:

This is really interesting. It leaves me wondering if the alignment of sectors really results in a measurable performance improvement in a typical Vista setup, or if perhaps we're worrying about nothing. I know the article claims a 50% improvement but under what circumstances?

One test that would be relevant to TrueImage would be to start with a clean Vista installation that is set up by the Vista installer to be aligned on 2048 sector boundaries. Then run a performance benchmark on the disk I/O; perhaps with SiSoft Sandra.

Next save an image of this partition and restore it using TI, assuming TI puts the start of the partition on sector 63. Then repeat the disk I/O benchmark and compare.

The last time I benchmarked my machine (on Windows XP) I measured about 60 Megabytes per second disk I/O using SATA 7200 rpm drives. Surely they're not claiming that this can be improved 50% to 90 MB/s??

 

I don't think the new sector alignment will improve the performance of existing systems. Rather, it's designed to support the soon-to-be-released "large-sector hard drives", and that's where the so-called performance increase will be coming from. Although these drives will offer somewhat improved performance, they will end up taking a performance hit if the sector alignment is incorrect. At least, that's my understanding.

snip...
"The file system, the volume manager, and other parts of the storage stack in Windows Vista have been updated to accommodate hard disk drives that have a large sector size. In earlier versions of Windows, the default starting offset for the first partition on a hard disk drive was sector 0x3F. Because this starting offset was an odd number, it could cause performance issues on large-sector drives because of misalignment between the partition and the physical sectors. In Windows Vista, the default starting offset will generally be sector 0x800."

Here's the link:
http://support.microsoft.com/kb/923332

 

Well I’ve ran dozens of tests with several different drive sizes on a couple of machines and the offset and ending sector numbers vary. I can’t see any pattern and have not tried to work out what the rule is. Vista will always end a partition before the CHS cyl boundary and I’ve seen this vary by several thousand sectors. When creating a partition behind one that does end on a cyl boundary then the new Vista one will always leave free space from hundreds to a couple of thousand sectors. The one constant is that if a partition is Vista “Cache Aligned” it will always end with the sector numbers xxxxxx2047.

The big surprise is that I discovered XP is not locked into cyl boundaries as I believed. If it creates a partition after a Vista one that is not on the cyl boundary, then the XP created partition starts on the very next sector – which is always xxxxxx2048. XP does however always seem to end the partition on a cyl boundary.

 

I have been reading this Thread, and not to throw another "monkey wrench", but you should read what I posted in my own thread:

https://www.wilderssecurity.com/showthread.php?t=175449

I am suddenly having a hard time making a CLONE of my drive, and best of all, the MBR gets corrupted on the SOURCE drive after trying to run the CLONE.

 

Hi Mystiky – I’ve answered in your thread.

To add to what I said above. Seems the Vista partition ending sector is not absolutely fixed on xxxxx2047 because yesterday I saw one ending on xxxxx4047. Also, it’s only XP’s Disk Management tool that will create a partition aligned with the end of a Vista one. When creating the partition during install then XP does seem to adhere to CHS cyl boundaries.

 

Your page MudCrab.
http://www.multibooters.co.uk/partitions.html

 

McTavish,

Thanks for the article. It was a nice read and explains the situation well. Hopefully all the kinks will get worked out soon.

I am curious though, if during your research, you found out why Microsoft decided to apply this new structure on today's current drives (with 512 byte sectors). As far as I can see, it's not helpful and is only causing problems. After all, Vista is not formatting the current drives with 4,096 byte sectors. On the "new" drives with 4,096 byte sectors it makes more sense.

If Microsoft had been more moderate, True Image, Disk Director and other programs that work with partitions would have had more time to get the necessary changes made to support it and the current programs would still work during the transition period.

 

Hi McTavish

A good lucid read and certainly one to add to my own knowledgebase. Tks

MudCrab in his last post was hoping that... "...all the kinks will get worked out soon..." But the more I understand about this issue the less confident I am that these so-called "kinks" can be worked out any time soon.

To quote from your article... "...The possible variations of OSes and partition types on multiple hard drives is going to be a minefield..." is probably about the most accurate assessment of the issue to date.

 

Hi MudCrab, I was asking myself the same question all through my writing of the article. Seems premature of them to make the new partitions the default standard. The only answer running round my head was the cynical one of that it will cut down on multibooting and cloning. It’s probably going to be a couple of years before these new drives are generally available and then quite a few more before they become the norm. We will probably be into the next generation of Windows before all the hardware and software catches up with the new drives.

Tabvla….Some of the third party vendors might crack it, others will have to just offer less features where Vista is concerned. I don’t think it’s insurmountable, just difficult. Personally I don’t like the automatic point and click apps that are aimed at the novice, as they get it wrong on too many occasions. Much better to know what you are doing and so know when you are trying to do something that is risky or impossible.

There will no doubt be an update to either XP's Disk Management utility or Vista’s, or both. I don’t do the update thing myself, just the service packs, so if anyone spots some MS patches in this area then I’d appreciate a heads up.

Cheers Guys.

 

I’m not sure that this post will add anything new to the thread but it will certainly corroborate a lot of the points already made. My hard disk as supplied by the manufacturer less than a month ago is divided into two partitions. One small 6GB hidden EISA Configuration used for Windows Recovery and a 292GB NTFS OS_INSTALL C:\ partition.

1. My problem started when I used Acronis True Image 10 (Build 4942) restore just my Vista OS_INSTALL C:\ partition. At first I thought it had been successful but then as Windows started to boot I got a BSOD (not had one of those since Win 95 days!) a few seconds after the initial Microsoft Corporation progress bar. It had the following Error Message:

STOP: c0000218 {Registry File Failure}
The registry cannot load the hive (file):
\SystemRoot\System32\Config\SECURITY
or its log or alternate.
It is corrupt, absent, or not writable.

2. I tried to reboot again while pressing <F8> to get the Startup Repair option to no avail.

3. I was advised by Acronis Support to do a Startup Repair but my OEM Vista installation disk does not include the Startup Repair option! I have since learned that this facility is also included on the Windows Anytime Upgrade DVD which is usually included with the purchase of a new computer. Has anyone tried it?!

4. After many anguished hours of hard work I eventually solved my own problem by restoring the entire disk i.e. both partitions + MBR. I later discovered this solution in the 34 pages of the “Vista: Are there issues with the disk ID?” thread on the forum. But it should not be necessary to wade through reams of often very technical posts to find a solution. By the way, I wondered if somebody could produce a summary of the practical recommendations arising from the discussion?!

5. Incidentally, Windows System Restore Points made prior to this incident no longer work. This makes me feel uneasy: can I trust Acronis to restore my PC if/when I have problems in the future? Is my registry corrupted? Are there still problems lurking beneath the surface of this apparently successful restore? Are there yet to be discovered consequences of this incident?

6. I have great respect for Acronis. They have a well earned reputation and are certainly more supportive of customers than most software companies. This is why I feel so let down. Previously I was a satisfied user of Acronis 8 which served my old XP computer very well for 3 years but after this experience on a new computer less than a month old I’ve lost confidence in the product. I wonder how many other users are merrily making Acronis backups every week thinking they are safe when in fact months down the line they will be in for shock! I feel I have been sold Acronis 10 under false pretenses: it is not fully Vista compatible! This issue is seriously damaging the other otherwise well deserved reputation of Acronis.

7. The more expensive editions of Vista (Ultimate, Business and Enterprise) include an image based “Windows Complete PC Backup & Restore”. It may not have all the Acronis bells and whistles but if it is reliable then an upgrade may be worth the cost.

8. In the automated email response to requests for help we are always told: “You can also quickly find the answers to your questions in Acronis Support Knowledge Base”. And yet there appears to be nothing in it about the need for running Windows Startup Repair after restoring a single partition on a multi-partitioned drive. Neither is there anything about it in the “User’s Guide”.

9. In his post to this thread on 8 May 2007, Aleksandr of Acronis Support said:

"Please be aware that there are several cases in which one should use Windows Vista DVD to perform Startup Repair after image restoration/cloning process. We are aware of this issue and Acronis Development Team is working on it.”

In that case why haven’t customers been alerted?! What’s the point of registering a product if the manufacturer doesn’t send you an email alert about a serious problem like this with detailed advice on what to do?! It’s the kind of issue that in other arenas would warrant a “product recall”! Failure to notify customers is irresponsible.

 

You should be OK.

This is yet another area where Microsoft has made some changes with Vista and the rest of the industry has not yet caught up. Vista uses a very low-level method (at the sector level) for creating restore points and shadow copies (previous versions of files reconstructed from saving sector maps of changed sectors). Programs like Acronis TrueImage do not put all of the sectors back into their former locations when restoring. The result is that restore points and shadow copies are lost after the image is restored. Even disk defragmentation software has to be careful not to move the restore point sectors around or they will also cause the loss of restore points and shadow copies.

I had a nice set of shadow copies going back over a month on my Data (D) partition that just got wiped out by running a boot-time defrag with PerfectDisk. I have since learned to exclude the System Volume Information folder on the drive from defragmentation in order to preserve the shadow copies.

To get to your original question, even though you lost the prior restore points, Vista will start creating new restore points and you should be OK. There should be no registry corruption or other problems caused by restoring your disk with your backup image.

Software vendors like Acronis have yet to fully grasp the implications of some of the changes made by Vista. Preserving system restore points and being able to keep the 2048 byte sector offset that is created by Vista's fdisk program are just two examples where problems have been introduced by not being fully aware of these changes. Hopefully they will eventually come to grips with these issues and fix them in future releases.

 

Thanks Mark. Your reassurance is much appreciated. I tried a System Restore and, as you say, it works ok.

There remains one last worry: my inability to perform a Windows Startup Repair. Pressing F8 as while booting up launches the Advanced Boot Options menu with its various options of which the first is “Repair your computer”. Selecting this is supposed to give me list of recovery tools including the desired Startup Repair. Instead I get a BSOD:

STOP: c0000218 {Registry File Failure}
The registry cannot load the hive (file):
\SystemRoot\System32\Config\SECURITY
or its log or alternate.
It is corrupt, absent, or not writable.

Does anyone know how I might regain this facility which resides on my hard disk rather then the OEM Vista installation DVD?

When I’m feeling brave enough (!) I might try accessing Startup Repair via the Windows Anytime Upgrade DVD which came with my new computer.

 

Hi,
I have read all these messages with much interest, because I had a problem when I restored an image of C: in my Vista notebook (Vaio CR120E). I purchased TI10 in July, installed, made images, made boot CD with Media Builder. Booting with Boot CD only works if I choose Safe Version. If I choose Full version, the computer hangs at the message "Loading, please wait...". Well, a couple of days ago I decided to recover an image of C: back to C: dated Aug. 1st. I started the process from within Windows, was asked to boot, then TI froze at "Loading, please wait...". Then, I booted from the boot CD, chose Safe version (because Full version doesn't work, I already tested), and the recover went to the end. However, adter finishing and booting to Windows Vista, I got the message: "Windows failed to start... etc. "winload.exe" is damaged (or corrupt, or missing). Repair your computer..." Very scared, I repaired with the recover DVD I made soon after I bought the notebook. So, there are problems with Vista, and I am considering a refund, because I can't trust TI anymore. And if I have the same problem next time I do a recover? And if the error is more serious and I can't repair? Will I lose my whole system or my whole HD wih all my work and information? I already contacted support, but no solution yet. So, 3 problems in hand: 1) Boot CD doesn't work for Full version (then, I cant use a USB HD, and I can't clone my HD to a bigger one); 2) Acronis hangs if recover is strated within Windoes Vista; 3) winload.exe is damaged and requires a repair if booting from the boot CD with Safe version. This is my testimony and I emphasize the words of someone, here: Acronis has to do something very quickly. Regars, Alexandre