You know, my gut tells me that this is a bad idea. There are a few things that make me uncomfortable.
Become a Patron of Ask Leo! and go ad-free!
Warranties don’t get your data back
First, realize what a lifetime warranty really means. If the SD card dies, you get a replacement for the card. But if the SD card dies, it’s taken whatever’s on it. Poof! All of the backups you may have stored on it are gone, and typically without any hope of recovery.
Now, of course, the same could be said of any external hard drive as well. They also die, though technically, the data stands a somewhat higher chance of recovery should that external drive be a traditional hard drive.
The difference here is what “lifetime” we expect from the hardware.
Flash memory: SD versus SSD
Now, I do want to make one clarification here. A $100, 256 GB SD card falls into the category of what I typically call “cheap flash memory”. And as we all know, flash memory wears out more quickly the more often you write to it. And periodic backups write a lot of data.
I personally would not trust my backups to a cheap SD card.
However, a $100, 256 GB SSD (Solid State Drive) is a different beast. Actual prices aren’t quite down to the $100 level, but I was actually surprised to see them getting closer.
While it technically uses flash memory, it’s of a far higher quality. It’s not a card; you can’t just insert it into an SD slot; it’s actually designed to be a hard disk drive replacement. You would install it in your machine or into an external drive enclosure – just like a hard drive.
SSD: still not for backups
While its lifespan is likely to be much longer than the cheap flash memory card that’s used in an SD card, I still wouldn’t use it for backups.
Why? Well, it’s a waste. The big advantage of Solid State Drives is their speed – specifically their reading speed. When used as a primary drive and a system drive, you’ll often notice significant improvements in your overall system feel in its speed. Many machines are now coming with SSDs as their system drives.
But that much speed you just don’t need for a backup for a variety of reasons. For the same money you can get at least twice as much traditional hard disk storage or even more.
SD cards are easy, it’s true
Now I realize the appeal of just being able to insert an SD card into the side of your laptop, for example, and use that for backup. And I’ll absolutely agree that it’s a 1000% better than nothing – but it has some inherent risks that I, again, personally would not be willing to accept.
Risks like the backup not being there when you need it.
Backing up is too important. Backup over the network to another machine if you don’t like attaching an external drive – or just bite the bullet and attach that external drive periodically.
And you know what? The latter is exactly what I do with my own laptop.
Great info here – thank you! I like to add that if one truly wants to back up their files, they should back up in at least two different locations. I accidentally deleted important files off the computer without noticing, and a month after that, the external back up hard drive failed. Had I backed up to a third location, say, a cloud drive, I would still have access to the files.
Agreed, JGM.
And with so many large capacity HDDs available, it is mainly remembering to do the multiple backups.
Additionally for my Family History files, these have been copied several times over the years, to DVDs and more recently, SDs, for distriubution to various relatives, so further protected.
I weekly backup my data files to a USB flash drive. I usually make one to keep at home and one I keep in another location, just in case a meteor or something destroys my house.
I have been using the same couple of USB flash drives for several years without any trouble. Leo, what is the live expectancy of USB flash drives (16MB USB 2.0) ? After a couple of years of weekly deleting and writing files should I think about buying a couple more and tossing the old ones?
That’s one problem. There is no real data on the life expectancy of flash memory as each brand would be very different from an other. Having a few copies in different flash drives should work. I once had a flash drive with a lifetime guarantee die on me. Sure, I got a new drive as a replacement, but if it was my only copy, the data loss could have been infinitely more expensive than the drive itself.
I thought I had the brilliant idea of making my one allowed copy of recovery media to an SSD card. I won’t be re-writing it, and, with any luck, I may never need to even read it. Right? Wrong? (What do I know?)
I can’t see the point. If you record a video (esp. of an important event), you are writing (if you do FHD or 4K), a LOT of data on that small micro SDXC card. Some do that all the time. They are made for this, some are specifically made for the massive volumes required for 4k recording. I can’t see the difference to writing an incremental image file to your card every day. Like, every 2 weeks a full backup (60 GB post compression), and every day some avg. 3 GB compressed for the incremental. Just as an example. A 200 GB micro sd card might — that is unless you can come up with a valid point other than gut feeling – fit this purpose very well – or not. The video analogy tells me it should. And the idea is, a micro sd card is so tiny, it can reside in your ultraportable laptop when you travel and you will always be doing your backups and incremental images. Of course you should remove it from direct contact to prevent troyans from getting access. since they are so small, you can even to a two card trick, so if a trojan or a thief gets his/its hands on your laptop and sdxc, you can laugh and say, well, I got the other one that I backuped to yesterday sitting in my purse or whereever you put it so you have control over it. (perhaps not your purse, but then again you take special care of your purse and if its just one of two, and encrypted, it might be not that bad an idea). And you are independent of mobile internet or your network or wifi issues. and seriously, IF it dies, you have that other stamp size sdxc card that has practically the same data on it. it is extremely unlikely that your hard drive and both of your sdxc die or get lost/stolen/infected simultaneously. You still should do an off-site backup whenever you get the chance, of course, and you will be moving data from the sdxc to an HDD once every 2 weeks of course.
Basically what we need is hard facts about cards like the sandisk ultra and extreme series of sd cards (both mirco and standard size).
I have not been doing this for long, but so far it has worked out.
btw. one might also be able tot use high quality usb sticks instead of an sd card. I personally trust sd cards more though. I usb sticks run a higher risk of being twisted when in your pockets etc. whilst an sd card is so small there are lots of ways of protecting it and still the space required for transport is minimal. I put it between two ATM cards sometimes… you can create your own safe space (sic) with some scotch and atm cards, or of course with a dedicated sdxc case. so if your laptop gets stolen or if you have to leave it unattended, you are on the safe side even in summer when you have little possiblities to carry stuff with you. Thinking of it, some might be able to put one of the two cards into their smart phones. though of course, bear in mind that if you are mugged, you might lose your purse and your phone and your laptop. so think again depending on how dangerous your current place is.
ps
http://www.tomshardware.com/news/sandisk-ultra-microsd-high-endurance,28657.html
“However, a $100, 256 GB SSD (Solid State Drive) is a different beast. Actual prices aren’t quite down to the $100 level, but I was actually surprised to see them getting closer.” That was written 2 years ago. Now 256GB SSDs are down to close to $100. But I got a 5TB external drive for not much more than that.
Years ago before hard drives existed, we always made 3 copies (1. a working copy) (2. a backup copy) and (3. a master copy) and we (a few of us taking great precaution) would ‘rotate’ the copies in a timely (before one was worn out… or upon a first small error, etc.) manner. We would make a new freshly formatted disk and copy and verify to it and it became the new (3. master) then we moved the old master to become the new backup, and the old backup to become the new working copy. We did this because the media itself became weaker just sitting in the drawer (critical data was in a fireproof safe). We added new disks as-needed when the data volume required to. I would not trust only 1 backup copy because it has happened where we lost one. As for the SD cards and chips and flash drives… those are scary to me. One mentions here using for a few years, I’ve had them quit working never to retrieve anything again… before the media was even filled one time. Once in less than 30 days and I treated them with great care etc. Maybe my experiences are rare but all that does is prove the point all-the-more to me that rare means ok…but! My worst experiences in a 3 month usage, was 1 of 3 flash drives trashed itself (couldn’t format, read, nothing) and 2 of 4 chips same thing in about 9 months. None of these storage devices was filled and deleted and refilled, all of them were never filled even once and one was only filled to about 1/3rd of its capacity. So I don’t trust them. Using hard drives over 30 years and numbers of them in the dozens and one pc running 7 years straight, I have never had one quit without warning me in time to recover (although I pushed it on one or two too far). Even doing customer repairs I’ve recovered data with the freezer-trick and others for customers, so the overall security I’ve experienced is sufficient. Most of the heavy stuff was done quite a while ago when programs like Norton Disk Doctor were in use more than you see today, and weak spots where data could yet be moved to a good spot and the bad/weak one marked as unavailable for future use… were good indicators for replacement if you let it get that far. I haven’t put pc’s through such tests in a long time and am not sure that the computers-and/or-the drive technologies warn people as well unless they also use a good utility to check once in a while. Doing so still is not 100% guarantee something can’t happen, but again I’ve never had it happen to me without some kind of warning first and with the memory media none of them has given me any warning at all, and I’ve never accessed one ever again that failed even though I’ve spent days trying 15-20 different utilities that claimed to recover data, none of them even knew there was media there.
I use USB flash memory for only 2 purposes. Transferring data from one computer to another and as bootable installation media. That would be for OS or backup recovery media which you could simply recreate if the drive is damaged. I would never use cheap flash memory for any kind of serious storage or backup for the reasons you mention here. High capacity USB flash drives can be a temptation to use for backup, but it’s not worth the risk unless you have at least 2 or 3 backups.
Thanks, useful info to think about. I have the unusual situation (probably) that I want to use MEGAsync to backup to my MEGA cloud storage, but this copies a designated local (or external drive) folder. It might seem fine to use the computer’s main drive (SSD), but it’s not very large, and I have a backup system that doesn’t compress files, and you shouldn’t have backups on your default drive anyway (even if it’s another partition, the drive could break). Hence an SD card struck me as a good intermediary, providing a backup itself on physical media that’s easy to store (or switch to another for ‘redundancy’). My own backup software writes the backup to the SD card, and then MEGAsync detects changes and updates the cloud storage. A complete failure of the card probably wouldn’t cause a problem, but corrupted files could possibly be synced to the cloud, making both copies bad – but that seems a slim possibility.
You didn’t respond to ‘micro sdxc card backup person’ above, and those do seem to be good points about SD cards for cameras dealing with a lot of data, but it will depend on how the technology actually works and how those backups are done (writing sequentially and leaving copies there, versus continually deleting and overwriting, perhaps – I’m no expert).
One thing I did discover is that some SD cards and USB flash drives – at least my old things – use FAT or FAT32 file system, rather than (Windows’) NTFS…although you can re-format them, so maybe this can be overcome (no expert, as I said). However, this can cause a MAJOR PROBLEM FOR BACKING UP if your backup program detects changed files by their modification time, which is often the case. FAT doesn’t store the full complement of a timestamp, meaning that the time recorded can be slightly wrong. This can cause programs that aren’t expecting the difference to miss files that need updating, or worse, to copy old versions over newer ones you already backed up.
I have a iMac using the macOS Mojave software. I have only used 100 GB of my Mac’s storage capacity. The macOS Catalina software is now available but I’m advised that I should backup my current software, files, photos, etc. before updating in the event that there are any issues with the upgrade.
Having only used 100 GB would it be OK to backup my entire used capacity (including my current Mojave software) to a 128 GB SD card or should I invest in an external hard drive?
The problem with an SD card, as the article states, is reliability. It should work for the backup but the risk of losing data is much higher.
I suspect the mac won’t let you. :-) But, indeed, I still recommend the safety and security of a “real” external hard drive. Let Time Machine do its thing to that.