How Windows Vista Backup uses zip files to store backups (and how to extract files from zips)

Published 04-10-2019 01:24 AM 297 Views
Not applicable
First published on TECHNET on Jun 19, 2007

Customers frequently ask us how to manually extract files from zip files created by Backup. Though using the Restore Files wizard is the preferred method, there are times when you must extract files manually, such as when you are using a computer that’s not running Windows Vista. I thought this would be a good opportunity to explain how the files get into zips in the first place, along with how to manually extract and combine files. I’ll also cover what file properties are not restored if you manually extract files instead of using the restore process.

(But before I begin, are you having trouble restoring files from a backup while using Windows Vista? See our previous blog post on the subject.)

If you’ve ever looked at your backup media, you no doubt noticed that each backup consists of a number of zip files, like so:

Instead of creating one giant zip file, Backup splits the files among a series of zip files with a non-configurable maximum size of 200 MB (or smaller if the media has less than 200 MB available). Backup is designed to avoid splitting files across zips and media if possible. The logic Backup uses is as follows:

·         Only files that are bigger than 200MB (before compression) will be split across one or more zips.

·         If a file is too big to fit a zip that already has other files, Backup will forcibly start a new zip, thus not splitting the file across zips if possible.

·         If the media has very little free space left, Backup will start a new media instead of splitting a file across media.

·         Backup delays backing up the big files until the end so that it can pick as many small files as possible before starting new media.

·         If a file’s streams are small enough to fit the current zip, but when combined they do not, Backup will start a new zip instead of putting streams in separate zips.  (Want to know more about NTFS streams? See Mark Russinovich’s tool called Streams .)

I will illustrate the steps to extract and combine a file using the following file as an example. Note the file size of the DISC1.MP4 file. You’ll see that after manual extraction and combination, the file is once again the same size.

So, let’s get started.

1.       I open my backup media (a DVD) and browse to the Backup Files <Date> folder. Here’s what mine looks like:

2.       I browse through the zips until I find my MP4 video, which is split into 5 files across 3 zips, like so:

3.       Since the files within the zip all have the same name, I can’t just extract them all to the same folder. I rename mine as I extract them as follows:

a. I create a target folder on my desktop named Humpty.

b. With the first zip file open, I copy the first file in the zip to the target folder. Then, I rename the file by appending _1. I repeat this for each file, incrementing the number as I go, including the last file I extract. Here’s what I end up with.

4.       I open the command prompt, navigate to the folder where my video files are stored, and then type the following command:

copy /b disc*.mp4 disc1.mp4

Here's what the output looks like in the command prompt window:






1 file(s) copied.

And here is the final result. Note that the assembled video is the same size as the one I backed up.

In cases where you need to extract a file that is split across zips or media, you need to be aware of what attributes and other properties are not preserved when you do so:

·         Alternate streams. The reason is that Backup stores them as separate files in the zip.

·         Timestamps for file create, modification, and access.

·         The file’s access control list (ACL).

·         The file path/name if the file path was longer that 260 characters at the time of the backup. (This is usually rare, since you can't touch such files from Explorer.)

·         Any characters from a language that doesn't have ANSI code page (aka Unicode-only language). For example, Hindi and Georgian.

·         Any characters from a language that is not the current language on the machine. For example, all Bulgarian letters on a Japanese machine. (This and the previous bullet are not Backup limitations, but zip format [designed for MS-DOS and 8088] limitations.)

·         The NTFS compression attribute. You’ll need to recompress the files after you extract them.

·         The sparse attribute.


Version history
Last update:
‎Apr 10 2019 01:24 AM
Updated by: