I do not think anyone has any doubts about the importance of our data and the need to protect them. Not just business data in your work, but the protection of your personal data should also be a constant concern.
Keeping in mind this concern of the importance of data that we back up every day, but did you imagine that just doing a backup is not enough? It's the old maxim of "who has one has none". Your backup may have been completed successfully, but will the file not have been corrupted for some reason? If your backup is recorded in LTO, will it be that when you need to restore you will not have problems with the tape?
Now, even making two backups, is it ideal to burn two LTO tapes? Two HDs? Two backups being sent to the same storage? In all these cases you may suffer from a generalized problem in your repository, if it is a single storage, and then those two backups you did with the concern of always having two copies of your data did nothing. What if the two LTOs you recorded gave trouble during the recording and did not you realize?
Lastly, even if you do two backups, it writes to different types of media, but leaves everything in the same physical location and a disaster happens (hurricane, fire, terrorist attack, you never know, right?), Again, all your concern.
What to do?
Thinking through all these scenarios I described that the 3-2-1 backup rule was created. Those who initially thought about this were Peter Krogh, a photographer, that is, a job that depends on the data and does not tolerate failures.
According to your idea, we need to follow three basic rules:
- 3 copies of your data (including the original copy, ie two backups);
- 2 different media types where your bakcups will be stored;
- 1 of these copies should be in an external location (cloud, site backup or tape storage locations, for example).
Now, with the concept and rule in mind, how can we implement it? Well, the big names in the data protection market offer you a way to follow the 3-2-1 rule. Let's, for example, exemplify with the Veeam.
Following the 3-2-1 rule with the Veeam
I will not explain a step by step how to follow the rule, but remember the options Veeam gives us at the time of making our backups. In the case of a file server backup, we would do the following:
- We would create our backup job from our file server pointing to a repository. This repository can be a storage or disk;
- Using the "Backup Copy" job we can copy the generated data to a second device, an LTO tape, for example and send it to an external storage location;
- If you do not use LTO, you could copy this data to the cloud using Veeam Cloud Connect which would also be an external location.
By following this example we would have three copies of the data (original, disk backup and cloud / LTO backup), we would be using two different types of device to store the backup (disk and LTO / Cloud) and sending the LTO to a location send your data to the cloud would ensure that at least one copy of the data is in an external location.
To conclude, I will leave some interesting links on the subject:
That's it, people. See you later!