Full backups. Types of backup. How to create a differential backup with Exiland Backup

A full backup contains all used blocks of data files.

An incremental backup of level 0 is equivalent to a full backup that has been marked as level 0.

A cumulative level 1 incremental backup contains only blocks that have changed since the last level 0 incremental backup.

A level 1 differential incremental backup contains only the blocks that have changed since the last incremental backup.

Full Backups

A full backup is different from a whole database backup. A full data file backup is a backup that includes every usable block of data in the file. RMAN copies all blocks to the backup set or image copy, skipping only those data file blocks that have never been used. For a full copy of the image, the entire contents of the file are reproduced exactly. A full backup cannot be part of an incremental strategy. Reserve copy; it cannot be the parent of subsequent incremental backups.

Incremental Backups

An incremental backup is either a level 0 backup, which includes every block in the data file except blocks that have never been used, or a level 1 backup, which includes only those blocks that have changed since the previous backup was taken. An incremental level 0 backup is physically identical to a full backup. The only difference is that a level 0 backup (as well as an image copy) can be used as the basis for a level 1 backup, but a full backup can never be used as the basis for a level 1 backup.

Incremental backups are defined using the INCREMENTAL keyword of the BACKUP command. You specify INCREMENTAL LEVEL .

RMAN can create tiered incremental backups in the form of the following RMAN backup types:

Differential: Default incremental backup type that backs up all blocks modified since the most recent incremental backup at either level 1 or level 0

Cumulative (Cumulative): Backs up all blocks changed since the most recent backup at level 0

Examples

To perform an incremental backup at level 0, use the following command:

To perform a cumulative incremental backup, use the following command:

RMAN> BACKUP INCREMENTAL LEVEL 1 CUMULATIVE DATABASE;

RMAN makes full backups by default if neither FULL nor INCREMENTAL is specified. Compressing unused blocks causes never-written blocks to be skipped when backing up to backup sets - even for full backups.

A full backup has no effect on subsequent incremental backups, and is not considered part of any incremental backup strategy, although a full image copy backup can be incrementally updated using incremental backups with the RECOVER command. This will be covered in one of the following articles.”

Note: You can perform any type of backup (full or incremental) of a database that is in NOARCHIVELOG mode - as long as the database is not open, of course. Note also that recovery is limited to the time of the last backup. The database can only be restored to the last committed transaction when the database is in ARCHIVELOG mode.

Various backup technologies have been developed over the years in an attempt to minimize the amount of disk space required to store backed up files and reduce the amount of throughput required to copy files to remote resources (computers, network drives and others). It is easy to get confused in the variety of backup methods offered by programs, as the terminology used is often not clear at first glance and does not describe the features of the methods. In addition, sometimes, at first glance, it is difficult to understand the advantages and disadvantages of any technology. This article is a guide that will allow you to understand some of the terms used, as well as their differences, advantages and disadvantages.

Note: This manual covers most of the main methods used today. Understanding their scope, limitations, features, advantages and disadvantages will be more than enough to make it easier for you to choose the right one for you. backup solutions on the computer.

General backup methods

Other backup methods and techniques

1. Full backup

This is exactly what it sounds like. This is a complete copy of all the data that the user selects when setting up a backup job. Usually, the copied files are placed in one archive file and compressed to reduce the size of the final backup. Every time a full backup is created, absolutely all files are copied from the source to the archive. There is one significant problem with this approach. Despite the fact that you have changed or added only a few files, each launch of the backup task will lead to a complete copy of the files, which will ultimately affect not only the duration of the operation, but also the disk space occupied, because each copy will contain a lot of duplicate files that do not differ from copy to copy. Of course, you can delete old copies to free up space, but it will still take a lot of time. In addition, if we are talking about storing backups on remote resources, then, in addition to time, the full copy will also affect the flowing traffic.

A much better idea would be to make a complete copy of the data once, and then only add or change individual files on a more regular basis. There are several methods that implement this idea and they are described below.

Advantages and disadvantages of creating full backups

Fast recovery of all files- When you need to restore a complete copy of files, it is easiest to do this with one archive file
Full backups take up a lot of space and take a long time- Full backups are not well suited for regular backups such as hourly or daily backups.

After creating an archive with a full backup, using differential backup helps to reduce the size of subsequent backups, making them based on a differential comparison of the original files with the files from the latest backup. All added or modified files are copied to a separate archive, next to the full copy. It is important to understand that differential backups are cumulative. Each differential backup backs up everything that is different since the last full backup, even if those files were already included in the previous differential backup. However, even with this limitation, differential backups are created much faster and take up less space than using the full backup method. Therefore, this method is well suited for daily or more frequent regular backup jobs.

Advantages and disadvantages of differential backup

Fast recovery compared to other methods- To fully restore all files from a backup, you only need two archives: the archive with the full copy and the last differential backup.
Differential backups take up more space than their counterparts- The approach used allows more efficient use of disk space and backs up faster than a full backup, but still this method still contains redundant data.
Each subsequent differential backup increases significantly- Since files are compared only with a full backup, sooner or later a differential backup will be compared with a full copy. In this case, you need to make a complete copy of the files again and start the process again (usually performed automatically).

The incremental backup method is very similar to differential backup, but has one fundamental difference - it stores less redundant data in backups. Each incremental backup contains only those files that have been created or changed since the last full backup or the last incremental backup. Such backups store a lot less redundant data than differential backups, but still the running total effect is still present, so incremental backups may contain files that were already in one of the incremental backups, but were later changed. Incremental backups are especially good for frequent backups, such as hourly.

Advantages and disadvantages of incremental backups

Incremental backups are faster than differential backups- By taking into account previous changes made, such backups store less redundant information and therefore are created much faster.
Incremental backups are smaller than differential backups- By taking into account the same previous changes, such backups store less information
You can create more incremental backups than differential backups- Since backups store less redundant information, there can be much more of them between full copies than in the case of differential copies.
Incremental backups take longer to restore than differential backups.- In order to restore files, you need to extract them from a full copy, and then successively apply all subsequent incremental backups.
Increased risk of information loss- If one of the incremental copies has been damaged or deleted, then restoring files from this copy will be impossible, as a result of which changes in files and added files will be irretrievably lost. However, it is still possible to restore data from other incremental backups.

The term "delta" is often referred to as a differential backup method, but is sometimes also referred to as "delta backup", "delta block backup", and "delta styled backup". And basically, all these concepts refer to the same backup technology. The delta method is most correctly called delta block backup, which is used in conjunction with incremental and differential approaches. It is important to note that the delta block backup method only applies to modified files, not created ones. The added files, of course, are also saved in copies, but in the normal mode.

The previously described backup methods create a complete copy of the changed file, even if only one character has changed in it. Of course, this approach will not be a problem when it comes to small text documents, but in the case of very large files, such as databases, this approach will be very problematic. For example, email clients such as Outlook most often store all information in one large file (letters, contacts, etc.). In this case, it turns out that even after receiving one letter, all previous methods will be forced to create a copy of the entire file. And since these types of files can change frequently, no matter what approach you take, your backups will grow in exorbitant steps and lead to the storage of a huge amount of redundant information.

Delta block backups deal with this problem by backing up only the parts of the files that have changed, rather than the entire file. The essence of the method is quite simple. Each file is broken into blocks of a certain size, and then the blocks of the changed file are compared with the blocks of the file in the full backup during the backup. And as a result, only those blocks that have been changed or added to the file will get into the backup copy. The term delta can mislead you, because depending on the methods used, the content in the created backups may be different. In the case of the differential method, the archive will contain the difference from the full copy, and in the case of the incremental method, the archive will contain the difference from the last archive with the modified file. Accordingly, the advantages and disadvantages will be the same as those of the methods in conjunction with which the delta is used. However, in the case of incremental backups, the risk of information loss will be higher, since the loss of an incremental backup will mean the inability to apply changes from all subsequent incremental backups (since it cannot be guaranteed that subsequent changes will be correctly applied).

Note: The block size will depend on the programs or user-selected size if the program supports it. Usually, the block size is in the range from 1 to 32 kilobytes.

Delta is particularly useful in technologies where files are backed up as soon as they are created or modified. This approach is also known as real-time redundancy or continuous data protection. Delta is also useful when backups are stored on remote resources (servers, storages) with limited bandwidth.

Advantages and disadvantages of delta block backup

Delta backups take up very little space and are created much faster
Delta backups allow you to store much less redundant information- Incremental and differential backup methods, due to the need to copy files, with minimal changes, can store a significant amount of redundant information. The delta block copy method can reduce this level.
Since delta blocks are created by programs according to specific algorithms, then you can restore them only with the same programs. In this sense, such backups will limit those users who may need to manually restore data.
Delta block redundancy is slower, as it is necessary to restore files from different parts.

Binary patch technology was originally developed as a way for developers to software it is easy to update your programs with customers via the Internet. Such "patches" replaced the modified parts in the files. Parts of files, not entire files. Such patches are much smaller than simple patches that completely replace files. Over time, this technology has been adapted to backup tasks.

Note: An example of such a backup method is FastBittm, which is used by large companies such as Microsoft, IBM, and Compaq.

The binary patch backup method is very similar to delta block backup, with the difference that delta uses blocks as the unit of comparison, while binary patches, as the name suggests, use bits as the unit of comparison. In other words, delta copies any changed data block to the backup archive, even if only a couple of characters have changed (for example, if the block is 32 KB, then even if 1 character is changed, the entire 32 KB block will be copied), and with the binary patch method, only changed bits are copied data. This difference allows you to save on size and, as a result, on the transmitted traffic.

Advantages and disadvantages of binary backup patches

Note A: There are currently very, very few backup applications that support this technology. In addition, there is very little information on this technology, so the disadvantages and advantages should be considered from the theoretical side.

Virtually eliminates data redundancy, resulting in the smallest backups.
The minimum size can significantly reduce the load to the network bandwidth.
This method is more demanding on system resources and time than delta. However, this approach can be offset by a significant reduction in network load.
There is practically no information about how the file is restored and how the method effective. Although, theoretically, the method should not be significantly more complicated than the delta, nevertheless, practice can seriously differ from theory.

6. Mirrored backups

Most backup software supports mirrored backups as an alternative to full backups, differential backups, and so on. Some programs use alternative terminology for the concept of a mirror, such as "simple copy". This is partly because mirror copies are basically a simple type of backup. This method does not use any special backup technologies, only a simple copy operation. As an example, if you copy and paste a directory from one drive to another, you can assume that you have created a mirrored backup of that folder. The files in mirror copies are usually the same files as in the source. They are not compressed into archives, as with a full backup (although some programs support individual file compression and encryption).

When to Use Mirrored Backups

Mirror copies without compression are well suited in cases where most of the copied files are already compressed into archives. For example, music files in mp3 or wma format, images in jpg or png format, video in DivX, mov or flv format. In addition, most installers are also compressed. If you include these files in a normal full backup procedure that uses compression, you will notice that in addition to the fact that such a backup will take a long time, the resulting archive will differ little in size (very little data will be compressed). In this sense, it is best to create separate backup jobs for compressed and uncompressed files. If your backup programs support filters, then you can use them to automatically select the appropriate files for each job.

Advantages and disadvantages of mirrored backups

Mirror copies are beneficial to use on compressed files.
Due to the fact that the mirror copy is not stored in a single archive file, you can worry less about file corruption.
Mirror copies generally do not use compression and therefore require plenty of storage space, unless other methods are used, such as hard links (more on that later)

7. Synthetic full backups

Synthetic full backup appears from time to time in the descriptions, but it should be understood that this is not a backup method, but a backup organization technology that can be applied to one of the above methods in order to restore and backup more efficiently.

Synthetic copies are usually used only in client-server backup systems. The meaning of the method is quite simple. The client computer can back up using any method (incremental, differential, etc.) and send this backup to the server. The server, at some point, will independently merge several separate archives to form a synthetic full copy. This way of organizing allows, after creating the first full backup, client computer Do not recreate full backups as this process is automatically performed on the server.

The advantage of this approach lies in two important points. Firstly, the backup speed when using differential copies will not decrease over time due to the accumulation of changes, since synthetic full archives will be created on the server on a regular basis. Secondly, the full restoration of files on the client machine does not require the process of reconstructing files from parts. The reason for this is that the reconstruction is already taking place on the server, allowing the client machine to restore the archives in the shortest possible time.

8. Backup using hard links

Some backup software supports the creation of hard links, which save disk space, when creating multiple full mirror backups of the same set of files.

To understand what a hard link is, you need to understand how files are stored on a hard drive. When a file is saved, the physical data can be written anywhere on the disk. In this case, the file system creates a hard link to the physical location of the data with the name of the file you used. Some filesystems allow you to create more than one hard link to real data. Using hard links allows you to use any number of files in different directories that will all point to the same physical data.

When using backup programs that support hard links to create multiple copies of identical files, the program will create hard links for all files that have not changed. For example, if you create two copies of a directory that contains 100 MB of data, then, under normal circumstances, these copies would take up 200 MB of hard disk space. With the help of hard links, such copies will take up the same 100 MB of disk space. Changing any of the files in such directories will actually change only the physical data, and the data will be available in both directories. For example, if after creating directories with hard links, you increase the file in the first directory by 2 MB, then their total size will be 102 MB, and the data in the file will be the same in both directories.

It should be noted that if you want to delete one of the backups containing hard links, then this will not be a problem, as the rest of the links are not affected. The physical data of a file on disk is only deleted when all hard links to it have been removed. It is also necessary to understand that hard links can only be created within one volume (logical disk). For example, hard links cannot be created between different partitions or disks. On Windows file systems, NTFS supports hard links while FAT does not.

Note: Windows Explorer does not take into account the use of hard links when calculating the size. This means that if a file is 100 MB and has two hard links, then in reality only 100 MB of disk will be consumed, while Windows Explorer will show 200 MB of disk usage. This point must be taken into account when using backup using hard links.

However, no matter how attractive this method may seem to you, it must be used with caution. Since the usual logic "changing files in different directories will mean changing different data" is not applicable in the case of this method. This means that an accidental change to one file in one of the copies will mean a change to the same file in all copies. Therefore, if you are using a backup program that supports hard links, it is recommended that you make all changes only through the program and refrain from making manual changes.

Final words about reservations

Despite the variety of methods cited, backup is one of those areas where the choice of technologies used should be justified in terms of the tasks to be solved. You should not use methods only because of individual characteristics, such as speed and disk space. So, for example, if you store backups on a portable hard drive, then using a technology such as delta block copy may save you space, but it can make such backups completely useless in situations where a backup program is not available ( for example, you need to correct the document on another computer, or for some reason the program has broken down, but there is no Internet access to download the installer). In this case, the use of conventional incremental copying (without delta) will be more justified, since you can always restore the necessary documents manually. Therefore, try to choose technologies with common sense.

Now, you know some backup terms, as well as understand what the methods mean in theory and in practice.

Incremental backup allows you to efficiently save information that is constantly changing: documents, projects in development, mail backup, etc. Handy Backup- incremental backup software any files.

What is incremental backup?

Incremental copy is a copying method in which additions are added step by step to the original copy of the data set, reflecting changes in the data (these step-by-step changes in the data set are called increments).

For example, if only 3 out of 200 files in the original set are changed, then they will be copied during the next incremental backup.

An incremental backup program should provide the ability to regularly make such additions, as well as incrementally restore them if necessary.

In this mode, incremental copying is repeated cyclically, starting from a full data backup. When the number of incremental backups reaches the specified value or statute of limitations, the whole cycle repeats automatically, starting with the creation of a new full copy.

The key advantage of a mixed incremental backup is its ability to break increment chains that are too long. This can slightly slow down the backup process, but makes data recovery more reliable and faster.

How to perform an incremental backup of files in Handy Backup?

Programming an incremental backup task in Handy Backup is very easy. Select in Step 4 in advanced mode* to create a task incremental or mixed incremental copying.

For mixed incremental copy also specify in the version settings dialog the length or time of repetition of the chain of incremental backups.

* At Step 1 of creating a task, you must check the box next to "Advanced mode".

Cross-platform incremental backup solution

Incremental copying of files and folders in Linux and over the network

In addition to the Windows version, Handy Backup also fully supports Linux distributions based on Ubuntu 16.04 and 14.04 at the executable level. The program also provides a workstation in Java for network Windows, Linux and FreeBSD machines.

Try the features of Handy Backup for incremental backup of files yourself,
by downloading and installing a free 30-day trial version of the program with all the features!

Today we will consider the principles of organizing incremental and differential backups using the program.

Often users use APBackup to completely save data, for example, to the same directory or to different archives each time using , as well as the archive depth parameter. This works well on small amounts of data. But if, for example, every day it is necessary to archive a completely large amount of information (for example, several tens of gigabytes), then a complete archive can take a long time, and slow down the computer. Although there is a mechanism that allows you to regulate the load on the computer processor (setting a low priority for the archiving process, automatic pauses in the archiving process, ..).

In this case, we will need to organize a backup using APBackup of only changed and new files since the last full backup, which will not take much time, especially in the case of backup to FTP.

What is the difference between incremental and differential backups? Let's say we made a full backup of the source directory and now every day we need to save changes to this directory. In the case of an incremental backup, every day the program will archive only new or changed files since the last backup (full or incremental). Thus, in order to restore the original directory in the event of a crash, we need a full backup and ALL incremental backups since the creation of this full backup. A differential backup will create an incremental backup of new and changed files every day since the full backup. Those. each next differential archive contains files included in all previous differential archives. When restoring, we only need a full archive and the LAST differential.

From the point of view of organizing backups, incremental and differential modes are similar, so in the future we will consider incremental mode in all examples.

So, for definiteness, let's say we need to organize a folder backup c:\work\ archive D:\backup\. We will do a full backup on Sundays (for example, a day off when no one is working with the server) and incremental backups every evening of the rest of the week.

The copy mode can be ANY, the program will work the same in any mode: Archiving (possibly using an external archiver), copying, copying to FTP. In our example, this will be archiving using an internal archiver.

So, for starters, let's create a task for organizing a full copy.

Let's name the task TEST_FULL, copy mode: "Archive", Backup type: "Keep all files"

Schedule: weekly on sundays.

A source: C:\WORK

To save the full archive, use the folder "d:\backup\", the archive has a prefix "FULL_" + date format. Depth = 1, i.e. only 1 last complete archive will be saved.

In principle, for reliability, you can copy the complete archive to additional directories on another server and even on FTP server in the same task.

Now that the full backup job is ready, you can create a copy of it to set up an incremental backup. A copy of the task can be made while in the main window of the program through the menu "Task"-> "Make a copy (F5)"

After the copy is created, the new job configuration window will open. We need to make the following changes in the new task:

Description: "TEST_INC", Backup type: "Only new and changed files (from the last archive)". This is just an incremental backup mode. To select the differential mode, you must select the copy mode: "Only new and changed files (since the last full archive)"

In the schedule, change the days of the week, select all days of the week except Sunday, when we will have a full backup

Bookmark "Saving archive" it is necessary to change the prefix of the archive to something different from that of the full copy, change it to "INC_". And also change the depth of the archive to 7 DAYS. Because to restore, we need ALL incremental backups since the full backup i.e. all copies in the last 7 days. In the case of differential copying, the depth can be set to 1 day, because we will need only the last archive.

Note that it is also possible to change the archive directory, for example, to save full backups on one disk and incremental ones on another.

After creating these two jobs, APBackup will work as intended, i.e. create a full backup on Sundays, and incremental backups on the rest of the week.