244618
4 Forms of Backup that System Administrators Should Know!
Backup comes in many forms. If used improperly, it will cause an impact on both the time required to restore data and wasted unnecessary backup file storage space, including the Bandwidth that we have to waste during the backup. Therefore, We would like to bring up an article from Networkworld that presents the concept of different types of backups to understand and use them correctly to get the most benefit.
Archive Bit is something you should know before making a backup. In Windows systems, Archive Bit is used to track whether files have been changed or not since the last backup. This will be cleared so that you know that they were last backed up and will not. It is backed up again until the next full backup is made (someone has a different opinion on why the name Backup Bit is not more accurate because it is not an archive).
But the problem is if we have more than 1 Backup application and are working at the same time. At the same time, there may be a competition between Archive Bits. However, in modern times we have used Virtualization This is done through the API at the block level, so the current Archive value is less important because it only looks at host-level backups. (The actual machine does not do this through the Hypervisor)
The 4 methods of Backup are as follows:
1. Full Backup
It’s a backup of every file, such as a Full Backup of Windows. Make a copy of every file on every drive like C:, D:, and others. If in Unix or Linux, start from /home, /opt, and others. The recommendation is You should only backup files that you know are not necessary, such as Config files like C:WindowsTEMP or /tmp in Linux. All of this is to avoid the case that other people’s Files that we don’t know may be placed elsewhere and may not have been backed up.
2. Incremental Backup
It is a backup of only data that has been changed from the last backup. Looking at the example picture below, it can be seen that the backup file will be small. This is the advantage of not interfering with the performance of the server much because it will back up a 10 GB file. Even though only 1 MB of data has been changed, that would be too much. But if you want to get Friday’s data, you have to take the data from the backup set from Monday to Thursday and put it together. If one is lost, there will be a problem. In addition, modern backup software currently uses Block-based incremental Backup, which is managed at the Block level by using APIs from VMware or Hyper-V.
3. Differential Backup
Backing up only the data that has been changed from a Full Backup, see the picture below. The popular method is to use a tape backup. To return the data, use a Full Backup and add the Backup file to the desired destination, which saves a lot of time than doing a Full Backup.
4. Forever-incremental
The arrival of the disk and Deduplication has made full and differential backups less popular because you no longer have to think about reducing the number of tapes. Including the time to restore data from a large amount of incremental data and a full backup, there is almost no difference because the backup system just keeps a record of where all the files or blocks are stored in Storage and just moves them back from Storage. It’s only on the client machine. However, the steps are not related to each other in the modern backup world, and the Forever incremental backup method will help with getting the most up-to-date data as well.
Source: networkworld.com
