I'm seeking a folder and file naming strategy for long-term backup management
I'm looking for practical advice on how to manage my backup files so that they'll be easy for me to manage and access over the coming 10 to 20 years. I'm concerned that long term backup file management will be difficult because Acronis gives all the backups for a drive the same file name. That means 10 years from now they'll all look the same. I'm not looking forward to having dozens of backup files all named Data1.tibx! :(
I manage an ever increasing amount of data, currently weighing in at 1.3 TB, going back more than 20 years. I anticipate wanting to maintain access to all that data, and more, over the coming 10 years. I want to be certain now that 10 years from now I will be able to find the backup for a particular date.
The data is coming from multiple people, multiple PCs, and multiple data drives on each PC. The backups are currently stored on around a dozen external drives, which is quite awkward. I can't have them all connected to my computer at the same time, and moving backups between drives is time consuming and tedious.
I start with keeping a full backup from each month, and then winnow that down over time. For the previous year I want all 12 backups. For backups that are 5 years old I keep 4 full backups a year. After 10 years, I keep just 1 backup per year.
Folder and file name conventions wouldn't be so much a problem if the Acronis user interface provided a full view of all your backups files and their contents, but whoever designed the current user interface sure liked white space, because there's lots of that, but multiple steps are required simply to see both file names and their contents. It's never all on the screen in front of you, and long file names are often truncated.
There used to be a clever hack that resulted in Acronis appending the backup date to backup file names. Because I used that in the past I can find and manage backups from as long as 10 years ago. I fear for the future now that I may have to rely on the modification dates of the files, and modification dates are too fragile to rely on, and tedious to view as well.
Currently the only I means I know for identifying backup files is to use folders that specify the date of the backups, so that's what I'm doing. Each month I create a new folder that includes the date and keep all my backups made that month there. I'm hoping someone has a better suggestion. Maybe there's a new way to include dates in the file names? Perhaps there are ways to rename backup files without confusing Acronis? Maybe someone has written a backup viewer?
Note that incremental and differential backup file names are not an issue, although I do have to worry about how changing file names might break my current backups.
One person I consulted on this told me that I should just use huge zip sets for long-term backup storage, rather than Acronis or any other backup system. Perhaps they're right? That would preclude future compatibility issues with new releases of Acronis and my very old backup files.
Anthony,
You ask great questions. I would say to you that your biggest enemy in long term storage of old data is that of data degradation, sometimes called data rot, data decay, or bit rot. If you store your data on magnetic media (HDD) for example, over time you will lose some data due to degradation. There are ways to prevent this but all require intervention of varying degrees. Because of this I cannot endorse the huge zip file suggestion. In general terms if you have data that is in the 10 year and older range, are you sure that data is still good? Have you tried accessing it? Do you have measures in place to mitigate data degradation?
As for naming, that aspect of data management is becoming increasingly more difficult with each passing year. For example, since the introduction by Acronis of the .tibx file format for full disk backup files, the backup method used to create these files are all dependent on each other which means that their names cannot be changed nor can their physical locations because if that does occur then the file is deemed corrupted and becomes unusable. This has created a situation where the user must develop a new approach to organizing, storing, and yes naming data uniquely.
Fortunately with the introduction of TI 2021 file renaming became possible again but carrying out that renaming is far less than intuitive. Another issue is that of backup Chain size followed by long term storage of Archive data. As for myself I have decided that all these factors mean that I needed to design a new backup strategy that could address these issues as well as a few others and would allow me a low participation rate in the management of my data itself. As a result I follow the strategy outlined below. It may well not fit your needs then again, you may find it useful by making some modifications that make if fit your needs. I hope it proves of benefit.
- Full disk backups: I only make full disk backups of my multiple machines System OS Drives. I make one to archive after an initial clean install of an OS with all available updates. I then create another full disk backup for archive after installing my current application preferences. These will be held until I have need to perform another clean OS install and normally span 6 to 8 years.
- Full disk backup scheme and method: For the above full disk backups I create them as Do not schedule - Custom scheme - Full method which means they are one off full backup files having no dependency on any other file. I like to say these are standalone backup files.
- User data backup: For my user data I now rely on Windows 10 File History and MS SyncToy 2.1 for backup purposes. I configure these two methods to make backup copies or syncs of only the user data I specify and to multiple locations. Since these are all data copies if I need a single file or folder of files I can easily navigate to one of the multiple storage locations and retrieve them.
So, I do not keep data that is decades old for a number of reasons such as advancements in file systems that make old data obsolete, storage space requirements, data degradation, and finally, over the years I found I simply do not have the need for anything more.
Having said the above, I do configure backups in the TI Windows application. Most are just for testing and really serve no purpose however, if I needed to get the data I could. I do run a single scheduled backup on each of my machines which I configure as follows:
- Schedule - Daily
- Scheme - Custom
- Method - Differential
- Options - Turn on Automatic cleanup. - Create 6 differential files before the next Full version. - Store no more than 3 recent version chains.
The above are my go to backups in the case of a drive failure or unrecoverable Windows OS failure. The above affords me with 3 full backup files and up to 24 differentials. I chose to move to the differential method instead of using the default incremental method because of file dependency. Differential files do have dependency with other files in the backup chain however, that dependency is simply that all files of the backup chain must be found in the same location on disk whereas incremental files carry this same dependency along with the fact that if any one incremental file of a backup chain becomes corrupted the entire chain is deemed as corrupted. This fact means to me that the incremental method of backup is not an option for backup when using TI as the backup product,
I hope that you can find a solution that addresses your needs.