All eggs in one basket
Being unhappy with both Windows 10 File History (no protection from Ransomware for backed up files) and the inability of the legacy Windows 7 backup program to do incremental/differential backups, I thought I'd try Acronis TIH's File Backup to backup important files to a folder dedicated purely for backups on my NAS drive.
The first backup seems to have completed successfully, but I now have a 1.7TB .tib file containing a backup of all my important files.
Instead of giving me peace of mind, I now have anxiety over this one great big monolithic file containing a copy of all my important files. After all, it's incredibly easy for large files to suffer corruption. All it would take is a small bit of corruption to potentially render my entire backup inaccessible/unrestorable. I also see no TIB repair tools in case it did get corrupted.
On top of that, I just double clicked the .tib file and it took ages to mount. Upon mounting, TIH promptly crashed and I can't browse the contents of the .tib file. Hardly instills me with confidence.
Correction, I can still browse the contents of the .tib in Windows Explorer, but it's painfully slow - taking minutes to open up just one folder. Hardly what I'd call usable.
On top of all that, I just went into the Recovery tab in ATIH and it says "No data to recover yet. Create your first backup version to see its contents here." Great!
I seriously expect more resilience from a dedicated back up program who's sole purpose in life is to safeguard important data. Why doesn't TIH create mutiple standalone TIB files? - each one holding a subset of the files and capable of being restored without relying on the integrity of the other files. That way a) each file would be a much more manageable size and b) if a TIB file were to become irrepairably corrupted, a much smaller number of files would get lost and not the whole bloomin lot.
Obviously there would be an index file with in-built redundancy to allow browsing/managing the complete backup set.
The current solution with everything in a giagantic .tib file seems very very flaky and unwieldy indeed.
And I did think about creating separate backup sets for different folders to achieve the same goal manually, but this just moves the problem to the user. It makes it harder to ensure all folders get backed up. And I'd have to remember to create new backup sets for new folders. Hardly fire and forget.
Have I missed something? Is this already possible? I've been a loyal TIH user since 2007 and have only really been using it for cloning and full OS disk images (typically 10-11GB max) til now.
3-2-1 is an industry minimum backup plan. 3 copies of your data, with at least 2 backups and at least one off-site and far enough away to survive the same event such as a fire or flood. Nothing is foolproof.
Regarding the large .tib.... The size of your data and backup depends on how you want to break it up. I purposely keep my OS and applications as one backup for fast and easy recovery. I have another backup for just pictures and home movies, another for my VMs, another for my music another for my software repository. This keeps back backup a reasonable size and separate from the others so if something happens, hopefully I can just recover just hat I need to and nit have to restore it all at once. But I can restore it all still if need be.
Secondly on the size... You don't have to create one large file. In the advanced tab, you can break the .tib files into smaller pieces. I have my backups set to 25GB to hopefully improve performance writing to disc. Just be aware that all pieces must be there to recover from.
That means if the disk you hold the backups gets corrupted, well, backups or pieces of it could too which would could cause trouble. This is why separating data a little and using at least a 3-2-1 plan is a must if your data is that important.
Thanks for the input guys. For some reason I didn't get notified of your replies even though the Notify option is ticked.
I take on board the suggestion for a 3-2-1 plan, but it does mean I will have to rejig some of my folder structure to make it work if I am to stick with Acronis for one of the backup tools.
I do think the single monolithic file is a chink in ATIH's armour, especially for file backups, more so given the larger file collections that are more prolific nowadays. I can live in hope that the devs will see it fit to make improvements in this area one day.
In the meantime, I guess I'll carry on using TIH for OS backups, cloning and perhaps backups of smaller collections of files.
Please note as posted above, you can set the size for your Tibs in the advanced tab of the backup scheme. Make them as small as you want, but you will need all of the pieces together in one folder, in order to recover from. I'm currently using the default setting of Blu-Ray so that my backup Tib files are broken into 25GB pieces instead of one large 100GB file.
Yes, I am aware the TIB file can be broken into smaller chunks, but as you point out, these are not stand-alone containers which is what I feel TIH would benefit from. As it stands, corruption in one container would still render the backup potentially useless.
That is true, but I have not seen that feature offered in any of the paid or free applications I've used and currently use (and I use just about all of them).
Keep in mind, your storage medium (hard drive) is a single container too (unless you're using multiple drives in RAID 1, 5 or 10, then you have a little wiggle room if one drive fails, but that's not common in most home environments without a specific NAS device - even then, corruption is mirrored to all the drives so it's still another "single basket" short of actual hardware failure of just one disk when RAID is implemented).
The solution is to break your data into smaller manageable chunks instead of putting them all into one big chunk. Do 1 backup for the OS, do another for your music, another for your pictures, etc. That way, if something happens to one type of data, you can deal with recovering it separately from all the other.
KB1972 wrote:I do think the single monolithic file is a chink in ATIH's armour, especially for file backups, more so given the larger file collections that are more prolific nowadays. I can live in hope that the devs will see it fit to make improvements in this area one day.
Perhaps I missed a solution somewhere in this thread, but I don't understand what you would like to see. Would you like a separate backup for every item (file or folder) you list you list in your backup task? I suppose you could submit a suggestion Feedback to Acronis or mention it in the Wishlist item in this forum.
If that scheme were implemented, I certainly hope it would be optional. It could result in a horrendously huge number of interestingly named backup files.
Patrick, the way I'm taking it is that he wants to be able to break up the .tib file, but also be able to pull whatever data is in each piece of the .tib directly out of those smaller chunks.
Seems like a nightmare scenario because there'd be no way to determine what files would be in each piece, if something like that was even possible.
Basically, you can break up the .tib file, but to get data out of it, it's still reliant on all the pieces being there in one place as they are not individual files, but just smaller chunks of the entire thing.
This is pretty much par for the course for anything that works this way. For instance, the same limitations exist in a multi-piece .zip, .rar, whatever file. It's the same behavior in all of the other backup products I've ever tried too.
It's still really just one backup file of whatever you told it to backup, but broken down into smaller chunks for different transport options. Having a missing, or damaged chunk, makes the entire .tib unusable though.
4th para from bottom in my original post:
"Obviously there would be an index file with in-built redundancy to allow browsing/managing the complete backup set."
Think of it like shopping bags. If you lose one bag, it shouldn't mean you lose your entire shopping load.
I don't care how other archiving/backup tools do it. I'm proposing something that's better than the status quo.
A simpler alternative to my original proposal would be something similar to Windows File History, i.e. files copied to backup location unmodified, but with TIH's ability to access the backup location as a different user.
Having seen ransomware encrypt all files in the Windows File History drive on a customer's machine highlights just how important it is to have access controls to this location. I can't believe MS are dumb enough not to think of these things.
I figured suggesting it to Acronis would yield a better chance of improving on the current situation. I will put it in as a suggestion once I figure out where that's done.
I appreciate hard drives are also a point of failure, but there are usually ways of recovering most if not all of a user's data from a dying drive. So the comparison to a .tib file isn't quite the same.
Roger that. I'm all for better solutions too! Also, submit feedback thorough the app and there is a current wishlist as a sticky on the 2019 forum. Can't hurt to add your thoughts in those places so developers are more likely to see them.
And some of what you're asking for does exist, but it is expensive in other products. Couple options below if you want to look into them...
Never thought of backups as shopping bags, more like a car that won't go if it's missing a critical component like the engine, brakes, steering wheel, transmission, engine, etc. Sure a few things can go missing and it might still work, but lose even just the key or the wrong fuse and your car isn't going anywhere, or it may go, but would you trust it on a road trip if you knew the engine block had a crack in it?
Unfortunately, the indexing becomes a database nightmare at this price point and local system resource hog that slows down the backup and recovery process. Most of the home backup software applications are working at the block level and don't keep track of the individual files so they can improve performance and keep the cost down with a fairly user friendly front end.
You're absolutely right that indexing is possible in backups.If you're willing to pay hundreds or thousands of dollars, there are Enterprise applications that offer indexing like what you're asking for. None of them I've worked with have the shopping bag method though and function properly when large chunks of the backup have been corrupted or deleted.
HOWEVER... Perhaps check out Retrospect. The server version I've used is a lot more expensive and I've had terrible experience with bare metal recovery and database corruption with it but it did have indexing, although still required all pieces of the backup to be present to recover from too. I'm not sure if the workstation version of it has the same options as the server versions but it's cheaper than the server (although about 3x Acronis though when I checked their website just now). Maybe check it out the trial to see if it has the features you're looking for at the higher price point? Another application that does indexing I'm aware of is Veeam Enterprise. I'm not as familiar with it but believe full backup chains are required to pull data from. It does have a nice indexing feature though. Starting price is 10 bundles at $960 though. I use their free desktop version for testing but it's really basic, can only backup one job and doesn't do any indexing or shopping bag like recovery.
Maybe one day, prices will come down for these types of features but probably not anytime soon. Just being realistic. If it does happen though, great!