Windows Server. Setting up software RAID

18.06.2018. How to protect Windows from problems caused by damaged hard disk? For this, there is a backup mechanism, in particular, one with the automatic creation of incremental or differential backups to maintain the current state of the OS. But there is also an alternative solution - mirroring Windows system partitions, performed by its own standard means.

This is the creation of software RAID in configuration RAID 1 in order to preserve information and gain access to the OS environment in case of problems caused by problems with the hard disk that ensures its existence. What are the features of this mechanism, and how to implement it in the Windows environment - we will talk about all this in detail below.

1. Windows mirroring: what is it

Mirroring- this, as mentioned, is a software RAID 1 , a commonly used disk array configuration in which data is duplicated onto a second hard disk called a mirror. In case of problems with the first, primary hard drive, we can use the mirror to access our valuable information. Moreover, if mirroring is applied to Windows system partitions, if the main disk fails, we will not only gain access to the information stored in the system, we will even get inside it. Not inside its original, but inside its exact clone on the disk-mirror.

Implementation of software RAID 1 possible in the context of dynamic disk technology. This technology exists in the Windows environment, starting from the version 2000 ... The technology itself is applicable to both MBR- and to GPT-disks, but here's the creation of software RAID 1 complicated by the need to perform additional operations with the command line. So everything that will be proposed below applies only to MBR-discs. Creation of software RAID possible only in OS editions, starting with Pro.

When reinstalling the system on dynamic disks, you do not need to inject specific drivers into the distribution RAID-controller, as required by the hardware RAID... As well as there is no need to reinstall anything when using any of the configurations of the software RAID... However, in the conditions of working with dynamic disks, we will not be able to use more than one Windows. The OS installed on other partitions will simply not boot. Technology works by rule "Entry - one ruble, exit - two": initial basic disks with structure and data are converted into a dynamic type easily and simply by Windows tools, but the opposite direction works only for disks with an unallocated area. If the structure and data are there, you will have to resort to third-party software.

Another important nuance: to work with this technology, it is important that the computer name contains only Latin characters. Otherwise, we will get an error "Invalid package name".

2. Preparatory stage

For Windows application RAID 1 a second hard disk with a capacity of at least the total volume of both system partitions must be connected to the computer. In our case, those occupy, respectively, 549 Mb and 60 GB, and the mirror disk has a volume with a small margin - 70 GB... The mirror must be prepared for its further fate - to delete all sections on it. There should be a clean unallocated area.

Information about booting installed on other Windows partitions, if any, is better to remove and leave the ability to start only the current system. When mirroring, the boot menu will be overwritten, and there will be a record of loading only one OS with the addition of the ability to run its clone on a mirrored disk. So it is important that there is a record of loading the necessary Windows. Otherwise, we get BSOD .

We will implement mirrored Windows using the system utility diskmgmt.msc, she's a console "Disk Management".

3. Convert disk to dynamic

On any of the two disks, call the context menu, select the conversion to dynamic type.

Check both of them. Click "OK".

We press "Convert" and confirm the action.

4. Creating mirrors of system partitions

So, both disks - both the main and the mirror - are now dynamic. We call the context menu on a small technical section of the system (bootloader section) ... We choose "Add a mirror".

Click on the mirror disk. We press Add Mirrored Volume.

Then we will see how a clone partition was formed on the mirror and the data synchronization process started.

Now we press the context menu on the main Windows partition, on the disk WITH ... And we do the same operation as above. Add a mirror.

Now we have configured data synchronization with the mirror. We can dispose of the unallocated area remaining on the mirror disk in any way: leave it as it is, create a separate partition from it, attach the space to some other partition (and on any of the disks, because now we are working with their dynamic type) .

5. Mirror Windows

As soon as the data is synchronized with the mirror, and we find out about the completion of this process by the degree of load on the disk in the task manager, we can reboot and test the performance of the mirror Windows. Access to it, as mentioned, will appear in the bootloader menu, it will appear with the inscription "Windows such and such version is a secondary plex" ... The bootloader menu, by the way, in the last two versions of the OS can be configured right at the stage of starting the computer.

You can set a shorter time for Windows autoselection.

The system will boot first on the main disk, so you can choose the minimum 5 seconds to display boot options.

In older versions of Windows, the timeout for the bootloader menu is configured in the system utility "System configuration".

6. Removing Windows mirrors

If you no longer need mirroring Windows, you can delete its mirror. This is done in the same place where this mirror was added - in the utility diskmgmt.msc... In turn, click each of the system sections, in the context menu, click "Remove mirror".

Select the mirror disk, press the delete button and confirm.

The space of the mirrored disk will turn into unallocated area, and its type is converted from dynamic to original base.

7. Reinstall Windows in a mirroring environment

Reinstalling Windows in the conditions of existence of mirrors of its partitions is carried out in the same way as usual - we can delete two of its partitions and specify an unallocated area with the OS installation location, or we can simply format its two existing partitions.

In any of these cases, when you reinstall Windows, its mirror will not go anywhere, it will continue to function in the new system. Program RAID 1 transferred to a new, reinstalled system environment. And everything would be fine if we mirrored regular user sections with our data. But the software RAID 1 for system partitions, remember, it also provides for the ability to enter Windows on a disk-mirror. And here we come across another Microsoft jamb: the record about loading the mirrored system will be lost - the very same bootloader menu item with an additional note "Secondary plex" ... After all, we formatted or deleted a small bootloader partition when installing the system. Leaving it as it is, not formatting it is even more dangerous. Let's remember that only one Windows can boot on dynamic disks. If the bootloader partition is not formatted, the new system will appear second in the boot menu and will not be able to start. Likewise, neither the first system nor its mirror will start, because the first one no longer exists, and its mirror is a clone of a non-existent Windows.

So the Windows boot partition when reinstalling it must be either formatted or deleted. How, then, do you ensure you log on to mirrored Windows? The solution here is very simple: you need to recreate the mirrors of the system partitions - delete them, as discussed in the previous paragraph, and reassign. The mirror disk is re-synchronized with the system partitions, and in the Windows bootloader menu, the item of the mirror system with an additional note will appear again "Secondary plex".

If you like articles, notes and other interesting material presented on the White Windows website and you have an irresistible desire to support this modest project, then choose one of two types of support strategies on a special page - Page with donation

Features of Software RAID in Windows Server

The following arrays are supported:

striped volume (RAID0)
mirrored volume (RAID1)
RAID5 volume
spanned volume (one logical volume is located on more than one physical disk)

Dynamic disks

RAID arrays can only be created on dynamic disks - a special layout of physical disks (understandable only by Windows), which has the following features:

A normal (basic) disk can only be converted to dynamic as a whole.
Converting a dynamic disk back to basic is possible, but only if you delete all volumes from the dynamic disk.
A dynamic disk is a single large NTFS partition, on which a large number of volumes (both simple and RAID) can be placed with the help of tricky service information, it is possible to resize simple volumes using standard Windows tools. However, I don't know how optimally and fragmented the data will be.
The cloning, repairing and resizing programs I know of do not support dynamic disks.
Dynamic disks containing RAID volumes can be migrated to another Windows Server computer as they contain the information required to properly assemble the array.

Unable to create volumes with different RAID levels

Only one type (level) of RAID volumes can be created on one physical disk group. For example, if we have 3 physical disks and we created a RAID5 volume on them without taking up all the space. We will not be able to create volumes of other RAID levels (RAID0 and RAID1) in free space, but only RAID5 and simple volumes.

Simultaneous volume synchronization

If several RAID volumes have been created on the same Disk Group, then in the event of any failure after the computer boots, they begin to recover simultaneously. This is a fierce, furious EPIC FAIL! A simple situation: there are two physical disks, two RAID1 volumes are created on them, one for the operating system, the other for data.

This scheme works great until the first failure (the simplest types are a sudden power outage or blue screen). And then the horror comes. The operating system boots up and starts synchronizing both RAID1 volumes at the same time. Thus, physical disks receive competing instructions for intensive sequential operations in three different physical areas at once. At the same time, the mechanics of the disks wear out wildly, the cache is useless.

From the outside, such "fault tolerance" looks like this: the overall speed of the disk subsystem drops every 20, the OS itself will boot either after the synchronization of one of the volumes (15 minutes, if it is small, gig by 50), or after 20 minutes and will be useless until end of synchronization of one of the volumes.

I consider the above behavior to be an unacceptable architectural miscalculation on the part of Microsoft and am surprised that this problem has not been resolved until now since the advent of software RAID in Windows 2000 Server.

If you find yourself in the described situation, then you should not wait for the OS to load and wear out the disks.

Disconnect one of the physical disks.
Boot into the OS at normal speed.
Break up the mirror by making the RAID1 volumes simple.
Connect the second drive back.
Create only one mirror for the volume you need.

RAID5

I will describe a scenario in which you cannot restore a degraded RAID5 array to a healthy state, even if all the conditions are provided for this.

There is a six-disk RAID5 array (Disk1-Disk6).
Among them there is one faulty Disk1 (for example, a couple of megabytes cannot be read from a terabyte volume), but the operating system does not yet know about it and has not marked it as faulty.
For some reason, a healthy Disk2 was disconnected from the array.
Following the logic of RAID5, if one disk fails, the performance of the array is preserved, such an array is marked as degraded, its performance drops sharply, synchronization with a new serviceable disk is required.
A working Disk2 is connected in place. The system recognizes him as faulty. To sync the array, this failed drive must be removed from the RAID5 array and identified as empty.
Everything is ready to synchronize the array. Run the repair of the array (repair) on an empty Disk2.
SUDDENLY synchronization stumbles upon read errors on a truly faulty disk, Disk1, and stops.
The entire massif remains degraded. Disk1 is marked as containing errors, Disk2 is marked as online, but due to interrupted synchronization it does not contain complete correct data.
In the hope of recovery, a brand new, working Disk7 is connected. Restoration of the array to it is started.
As a result, the healthy Disk2 is replaced with another healthy Disk7, but the synchronization is interrupted again, having found an error on the defective Disk1.
And so on in a cycle.

You cannot do anything except copy the data that is still readable and rebuild the entire array.

Not wanting to admit defeat, I tried to do the following things:

Synchronize the array with skipping read errors on the Disk1 disk (after all, these are just megabytes of their whole terabyte). But Microsoft doesn't offer that option.
Overwrite the entire faulty Disk1 sector-by-sector to another healthy disk using cloning programs. However, the programs available to me did not work with dynamic disks.

An example of a competent implementation of software RAID

The hardware and software implementation of the RAID controller, known as Intel Matrix Storage, and recently renamed to Intel Rapid Storage (runs on RAID versions of chipsets such as ICH9R, ICH10R), has been eliminated from the above drawbacks. Intel hardware / software RAID provides many of the benefits of an "adult" RAID controller:

ability to identify hot swap drives
the ability to create volumes of different RAID levels on the same disk group
sequential synchronization and verification of RAID volumes on a disk group

Its main drawback, in contrast to fully hardware RAID controllers, is its "software" nature, from which follows:

lack of built-in cache and the possibility of autonomous operation in case of an accident
depends entirely on the operating system and drivers
the operations performed on the disk subsystem load the main processor and memory
no support for advanced computationally intensive RAID levels such as RAID6

useful links

What Are Dynamic Disks - Windows IT Pro [fairly old article]
The Whole Truth About Dynamic Disks - Hacker [read carefully, "all" truth mixed with fables]

Setting up a software RAID array under Windows is much easier than under Linux systems, but it also has its own peculiarities. Often, incomplete and fragmentary knowledge in this area leads to difficulties, and among administrators there are myths and legends about the "capriciousness" and "glitchiness" of this mechanism in Windows. In this article, we will try to fill this gap.
Before continuing, let us again recall the basic principle of building hardware arrays: one array element - one physical disk. The basis of software arrays is a logical disk. Understanding this difference is the key to success, what is applicable to a hardware array can be disastrous for a software one, especially when it comes to the failure of one of the array elements.

To create software RAID in a Windows environment, we need to get acquainted with the concept of a dynamic disk, since software arrays can only be created on them. The reputation of dynamic disks is ambiguous, many administrators shy away from them like hell with incense. And in vain, remembering a few simple rules, working with dynamic disks becomes as simple as with ordinary ones.

As a general rule, installing or booting Windows from a dynamic volume is only possible if this disk has been converted from a system or boot volume. Those. if you have several copies of the OS, then after converting the disk to dynamic, you can only boot the instance that is on the boot partition.

Based on this rule, it becomes obvious that only a mirrored array (RAID1) can be created for the boot and system volumes, it is impossible to create other types of arrays, since they imply the installation of the system on a previously created partition.

Is it worth the candle? Despite all the limitations, it is worth it. The main disadvantage of hardware arrays is the binding to a specific controller model. If your motherboard or controller burns out, you will need exactly the same one (or a motherboard with a similar controller), otherwise you can say goodbye to data. In the case of software RAID, a machine with Windows Server installed is sufficient.

In practice, work with software arrays and dynamic disks is done through a snap-in Storage - Disk Management v Server Manager... To convert disks to dynamic, just right-click one of them and select Convert to dynamic disk, in the window that opens, you can select several disks to convert at once.

It is worth remembering that this operation is irreversible and special attention should be paid to the system partition, you will not be able to repartition the boot disk (or rather, it will cease to be bootable after that), the only thing you can do is expand the volume using unallocated space.

The next step is to create an array, right-click on the desired volume and select the desired option, in the case of system and boot volumes, there will be one option - a mirror, then you will be prompted to select a disk for placing the mirrored volume. Upon completion of the creation of the array, its resynchronization will immediately begin.

By connecting additional disks, we will get much wider possibilities, you can either combine several disks into a separate volume, or create RAID 0, 1 or 5.

In general, nothing complicated, but many restrictions can scare anyone away. But do not rush to make hasty conclusions, judging by common sense, there are no serious obstacles, since it is usually customary to distribute the system and data across different disks, given the penny cost of modern disks, this does not entail significant costs. For example, we created a mirror for the system disk and RAID5 for the data for our test server.

Moreover, all this pleasure can be realized on the most ordinary budget motherboard, given that the performance of a software array is no different from cheap hardware, this technology looks very attractive. We will talk about methods of ensuring fault tolerance and actions in case of disk failure in our next article.

How to create a software RAID 1 array (mirror) using Windows 10, 8.1, 8, 7, Vista.
What is RAID 1?
RAID 1 is an array of two disk media, information on which is duplicated on both disks. That is, you have two disks that are full copies of each other. Why is this done? First of all, to increase the reliability of information storage. Since the probability of failure of both disks at the same time is small, in the event of failure of one disk, you will always have a copy of all information on the second.

You can store any information on a RAID 1 array just like on a regular hard disk, which allows you not to worry about an important project that you have been working on for a very long time.

Today we will look at how a RAID array is created using Windows itself when using two empty disks (I confidently declare that this instruction works on Windows 7, 8 and 8.1).

Instructions:
1) First, install the hard drives in the system unit and start the computer.

2) Open “Control Panel → System and Security → Administrative Tools → Computer Management → Storage → Disk Management”. When you turn it on for the first time, the utility will inform you about the installation of new disk devices and offer to choose a partition for them. If you have a disk of 2.2 TB or more, choose GPT, if less, then choose MBR.

3) At the bottom of the window, find one of our new hard drives and right-click on it. We select "Create mirrored volume":

4) The Image Wizard will open. Click further.

5) On this page you need to add a disk that will duplicate the previously selected disk. Therefore, select the disk on the left and click the "Add" button: Click next.

6) Select the letter that will be used to designate the new volume. I chose M for Mirror. Click on next.

7) Set the file system, cluster size and volume name. I also recommend checking the box next to "Quick Format", let it do everything at once. And again further.

8) We check what we have done, if everything is correct, we press "Finish".

9) The system will warn you that it will convert the disks to dynamic, and we will not be able to create a boot volume on them. There is nowhere to retreat, we press "Yes". :)

Read about the Windows Disk Space feature. Namely, how to create a software Raid from several physical disks on your home PC yourself, including built-in or connected via USB using this function, and restore data from it.

The Disk Spaces function built into Windows 10 allows the user to create one virtual hard disk from several hard disks. With its help, you can back up data by duplicating on multiple disks or combine several hard or SSD disks into a single storage pool. Disk Spaces is the same as RAID, only on a desktop computer and with hard drives attached to it.

Content:

Windows "Disk Space" feature

This feature first appeared in Windows 8 and has been improved in Windows 10. Storage Spaces is available in all versions of Windows 8 and 10, including Home.

To create Disk Space, at least two physical disks must be connected to the computer, including built-in or USB-connected.

Using the Storage Spaces function, you can create a storage pool of two or more physical disks by grouping them together. After creating a storage pool from two or more physical disks, you can create three resiliency spaces:

Simple type... Designed to create a disc of the largest possible volume. This type of space does not protect data in any way in the event of a possible failure or failure of one of the disks. On such a disk, Windows will save all data in one copy. If one of the physical disks fails, all data stored on it will be lost. This type of disk is convenient when you need to temporarily store large amounts of data.
Mirror space is designed to protect data in case of a possible failure of a physical disk, by saving multiple copies of files. In the event of failure of one of the physical disks of space, the data stored on it will remain accessible from the other disk on which their copy was created. This type is useful for protecting sensitive data in the event of possible hardware failures.
Even space Is a cross between a simple and a mirror type. Data is striped across multiple physical disks, creating one or two copies of the parity information. However, due to the need to calculate checksums, parity space is noticeably slower for writing, as a result of which they are recommended for storing data archives. For example, a photo or video.

How to create disk space

Create Disk space can be done using the corresponding menu in.

But before you start creating it, connect to the computer all hard drives from which you intend to create Disk space... Then select from the menu / Create new pool and disk space.

Select the disks you want to add to the pool and click the button "Create pool".

Please note that all disk data from which the disk space is created will be deleted.

After creating a pool, you need to set up a new disk space: assign a name to it and select a drive letter. It is with this name and letter that it will be displayed in Windows.

Also, you can choose the standard Windows file system - NTFS, or the new type of system ReFS. In the case of creating a mirrored or parity space, which are designed to protect data from loss, it is better to choose the type of ReFS system.

Indicate the type of resistance: Simple (no stability), Two way mirror, Three way mirror, Parity.

To create a large storage pool without disk failure protection, select the type Simple(no stability). A two-way mirror involves storing two copies of the data on disk, and Trilateral- three. Disk space with type Parity will protect in case of failure of one of the disks, and will be larger than a two- or three-way mirror, but will be much slower.

Depending on the selected resilience type, the wizard will set the maximum currently available disk space. However, you can also set a larger data pool size. This is intended for the case when the available space of the connected physical disks becomes full. So that the user can connect another disk without having to make changes to the disk space configuration.

Then press "Create Disk Space".

How to use Disk Spaces

The created disk space will appear as another disk in the folder "This Computer"... Such a disk will have the name and letter that you assigned to it during creation and will not visually differ from other disks.

You can do everything with this disc as with another regular disc. Even encrypt it with Bitlocker.

How to manage disk spaces

After creating disk space, you can go back to Control Panel to manage or customize it.

You can create another disk space. Their number is limited only by the number of physical disks connected to the computer. You can add disks or rename the storage pool.

You can change the name or letter of the disk space by clicking the menu "Change" in subsection.

To add disks to an existing disk space, select Add Disks and select the disks you want to add. By choosing the Optimize Disk Usage menu, Windows will redistribute existing data evenly across all disks.

If the existing disk space consists of three or more physical disks, then one of them can be removed. To do this, expand the physical disks menu and select the link "Prepare for deletion" next to the drive you want to remove. In our case, there is no such link, since the disk space consists of two disks.

After that, the system will transfer the data from the removed disk to two (or more) other physical disks of the disk space. As a result, the link "Prepare for deletion" will change to "Delete".

The physical disk removed from the disk space will become available in Disk manager... For the possibility of further work, you may need to create a new section on it and format it.

In the management of disk spaces, the user also has access to the function of deleting the disk space itself (to the right of the disk space name, subsection). Just keep in mind that all data in the deleted disk space will be deleted.

In case of deleting disk space, the menu will become available "Delete pool", by selecting which you can completely remove the storage pool.

How to recover data from a disk space or mirrored volume

What to do in case of data loss from disk space? Is it possible to recover them?

In the event of failure of one or more disks of disk space, the process of recovering data from it becomes very complicated and cannot be reduced to scanning each disk separately with a data recovery program. In this case, the data will not be recovered or will be recovered damaged.

The only exception is mirrored disk space, which is created as RAID-1. Since in a disk space of this type, a copy of the data is created on each of the disks.

To do this, it is enough to scan the logical disk in the form of which the disk space is displayed using Hetman Partition Recovery. Then find and recover lost files or folders in the same way as from any other disk.

How to create a mirrored, striped, or spanned volume in Windows 7 or older

As mentioned at the beginning of this article, the Disk Spaces feature has appeared in Windows since version 8. But before that, the system also provided the ability to create disk spaces. You can create a mirrored, striped or spanned volume in Windows 7 or older using the Disk Management menu.

Because Disk management is present in Windows 8/10, so you can create disk space in the latest versions of the system in the same way.

To create disk space using Disk Management, you need two or more disks connected to the computer, which will not be allocated. If a partition will be created on the disk from which you plan to create a mirrored volume, delete it. To do this, right-click on it and select "Delete volume ..."... Note that this will erase all data from the disk.

After that, right-click again on the disk from which the disk space is supposed to be created, and, depending on the task at hand, select the type of volume to be created. They are somewhat similar to the types of disk space resiliency described in the first part of this article.

So, Simple volume Is a part of a physical disk that functions as a separate physical block. It is nothing more than a normal logical partition on a physical disk. It is not disk space, as we are talking about in this article.

Composite volume connects areas of free space of two or more physical hard drives into one logical drive. It consists of at least two unallocated parts on two hard disks, which are combined into one common when created. If you create a spanned volume that includes 150 GB from one hard drive and 250 GB from another hard drive, a 400 GB local drive appears in the This PC folder. The performance of a spanned volume is higher than that of a simple volume, and the fault tolerance is the same. The creation of a spanned volume is suitable for solving the issue of the size of the local disk.

Striped volume is a non-redundant array configuration. Information is split into blocks of data and written to multiple disks at the same time. Failure of any disk destroys the entire array. Since the array is not redundant, there is no recovery procedure in the event of a single physical disk failure. Reliability depends on the reliability of each drive. The array is not intended to provide reliable storage of data, but to improve performance.

Mirrored volume Is the same as the mirror space. It is designed to protect data in case of a possible failure of a physical disk, by saving multiple copies of files. In the event of failure of one of the physical disks of space, the data stored on it will remain available from the other disk on which their copy was created. This type is useful for protecting sensitive data in the event of possible hardware failures.

Volume RAID-5, as well as even space, a cross between a simple and a mirror type. Data is striped across multiple physical disks, creating one or two copies of the parity information. However, due to the need to calculate checksums, parity space is noticeably slower for writing, as a result of which they are recommended for storing data archives. For example, a photo or video. To create it, you need three or more disks.

I choose " Create mirrored volume ... " and in the resulting Image Wizard, click "Further".

Select the disk that you want to add before the mirrored volume (other than the one that was originally right-clicked) and click "Add" / "Further".

Assign a drive letter.

Give the volume a name and format it.

Before formatting begins, the system will warn you that it will convert the basic disks selected for the mirrored volume to dynamic disks.

This will start formatting and converting the disks to a mirrored volume. In the Disk Management window, the disks of the mirrored volume will appear in brown, and in the This computer folder, as one of the local disks.