Would this work or would I have problems:
Using dd command to backup an entire SSD containing dual boot Windows/Ubuntu partitions into an .iso file, with the intent to then dd that iso back onto the same size SSD in the case of a drive failure?
Would this work
Yes.
or would I have problems
Also yes.
I used to do this backing up my “servers”. By that I mean some Raspberry Pis and random old PCs running Debian. I even did so successfully when needing to restore the images. But it was fragile and also failed at times, sometimes to great inconvenience when it was a machine serving something important.
I’ve since moved to a different backup strategy for servers, but if I were to do this with a bare-metal machine I want to preserve, I’d use something like Clonezilla. The maintainers of that project know a whole heck of a lot more than I do of the ins and outs of disk management, backup, and restoration than I do with my simple dd commands. If it is something you’re just wanting to do for fun and experience, dd can work. If you’re concerned with the security of your data/image, I’d use Clonezilla.
Honestly, that just seems like you’re treating dd as some kind of arcanum. dd works just fine and I’ve been doing 1:1, full system backups for decades with it, no issues. Honorary mention for ddrescue / dd_rescue for recovery options, i. e. re-trying bad sector reads etc.
In fact, when Clonezilla doesn’t know your filesystem, it will simply employ dd to copy the data sector by sector.
I’d argue that Clonezilla (due to its use of partclone) is actually a less complete form of backup, since it will only copy used blocks, you don’t really end up with a clone of your devices, just a copy of what partclone believes to be your data. Don’t get me wrong, that is fine in most use cases, but there are some cases where this doesn’t cut it, e. g. wanting to backup / restore a storage device from a PLC where the vendor had the glorious idea to store licensing data in unused sectors, or when you want to create a forensic disk image, might want it look into d3dd then, although it absolutely works using regular old dd as well, d3dd just adds some amenities.
All I want to say is: dd is an absolutely reliable tool and can be a one stop solution for device backups. Also, I have absolutely no quarrels with Clonezilla, if it fits what you’re trying to do and it works, great.
As you seem to know a lot about Clonezilla, I’m allowing myself to ask you something.
Would Clonezilla work for doing a backup and reinstalling it on a completely different device?
I’ve already used it to backup and restore on the same device. I also tried to install my Surface Go device backup on an old Acer Computer but it failed to boot afterwards. I’m not sure if it’s because of the surface specific kernel or because it doesn’t work from one computer to another.
Would you know that?
I don’t know off the top of my head. I think that Clonezilla can modify images in such a way as they can be booted on a different type of device. My knowledge of the black magic of boot sectors and partition stuff is lacking. Also, you’d have to make sure the motherboard/BIOS is properly configured for reading the device in the same way that the original device was read. UEFI/BIOS stuff can be a pain in the ass to get right.
So my short answer is probably, but I wouldn’t be able to walk you through something like that. Wish I could be more helpful.
In general, no, this won’t work. In your case, you’re lucky since at least the Surface Go is using an x86 CPU, so it’s not completely out of the question, but transferring the image as-is to a completely different device typically does not work without modification.
Simple example: your target device might not refer to existing hardware (let’s say a storage medium) in the same manner as your old device, so the existing references in your cloned image won’t work. There are other issues of course, e. g. missing drivers for different hardware present on the target device.
It’s possible to modify the image so it would boot, but given the Surface runs Windows, that’s going to be a chore. I’d consider this an interesting project if bored on a slow weekend, but I’d most likely just do a filesystem backup of relevant data and call it a day.
I forgot to say that I’m using Linux (Fedora 40). If you think it wouldn’t work, do you know another program which would?
My data is stored on the cloud, but I wouldn’t want to go through the hassle of setting everything up again if I move to another computer (extensions, mail, calendar, installed programs and settings).
As long as it’s not a mounted and running live system, it should work. Not the most efficient way to do such a thing, but I would think it should work.
You can zero out the free space on each partition then pipe the output from dd into gzip if you want to save space.
Don’t do this. If you have a 1TB drive with only 10GB being occupied, your image will be 1TB, and you will need a >1TB drive to store it, and another to restore it.
If you only backup the data you could do it in a much smaller size drive and it will be a lot faster to perform backups (otherwise you will need hours every time you want to create a new image).
Works fine, though it’s not an “iso” file. But it doesn’t matter what extension you use.
I used to do this to switch an old laptop between Windows and Linux. I’d backup one, overwrite with the other. Swap as necessary.
Yes. Make sure neither are mounted.
Use larger block sizes with bs=<like, idk, 16M> will make it go faster.
