Balena Etcher Pro – Fast write to multiple cards or USB disks at once
balena.io
balena.io
neat tool if this is actually a need. the expansion/cable management looks pretty slick.
on a related front, I do sometimes wish that I could burn a lot of CDs all at once, but not often enough that I am going to buy a $1000 piece of dedicated hardware. One at a time with ImgBurn it is!
That used to be the case when CD-based piracy was rampant and you had groups with literally 1000s of CD-writing burners to mass produce pirated copies of movies/software/music etc
Now the use-case for this Balena device would mainly cater to brands wishing to promote their product with a custom USB flash drive logo and some software/promotional material found on the drive
Just envisioning some uses off the top of my head.
At my previous job we’d have recouped the cost within a couple of months at the most, and the sanity of whichever person was assigned to do the cards.
manufacturers offer pre programming, its an added cost and might have MOQ requirements
your factory offers to build an automated programming jig, again makes sense at a certain volume level
you and a bunch of friends and family sit around and program the stuff manually for couple of days straight
For those complaining that it’s more expensive than cobbling together USB hubs and multi-port readers from Amazon: You’re definitely not the target market. If you have a scaled manufacturing or even prototyping facility, having an engineer spend a couple days cobbling together scripts and instructions to do the same thing as well as maintaining it on the floor will cost more than this device. On the other hand, if your a hobbyist with plenty of time but limited budget then this isn’t for you.
Then again, if I was wanting to burn 1000s of cards a day, I would probably do what everybody else does and create/contract an actual manufacturing process. This seems to fall in the niche between somebody that just happens to flash a lot of images and businesses at scale that do 1000s and it seems like it's a bit of a perilous position.
In practice, lead times are long enough and release cycles short enough that it can make sense to have these devices at point of manufacture so you can ship with the latest software image.
Imagine you order 10,000 pre-flashed SD cards for your project but half way through the production run you decide you need to change the software. Do you discard thousands of SD cards and re-order a new batch? Or do you buy a couple of these and simply add it as another station on the assembly line so you can get back to production right away?
No, you don't pull a run of hardware mid-run to rev their firmware unless there is an absolute shitstorm which would also just be handled by the manufacturing process i.e. why would you ship the "bad" cards to your office or some such scenario rather than run them again through the same at-scale manufacturing process?
Anyways, as somebody already pointed out, the vast majority of these cases would be handled by live updates.
EtcherPro uses the standard open source Etcher app so hobbyists can use USB hubs to build their own drive duplicator on a budget.
- EP lead designer
Why did they have to backdoor the thing? Now you don't know if it's scanning your content or extracting crypto keys and sending them to master control. It's a copier. You had one job.
Weird to see people taking the exact opposite stance and asking "who is it for".
Obviously this is _awesome_ for any large deployment of Pi's.
Imagine you'rea teacher and you're giving Pi's to kids (their intended use case), I wouldn't want to write 30 cards one by one.
Looking at the price, it's probably justified, but hard to switch when I have a working solution that was $23 + a couple hours of coding.
Again, I think this device isn't a terrible value, but a $2-300 device would be a lot sweeter proposition for my demographic.
If you are pumping out 100’s or 1000’s of these a week yet alone a day you are probably making money of it or supporting an activity that does.
And writing the cards can be done with Balena Etcher https://www.balena.io/etcher/
Works well enough and much cheaper than the Pro version. Using it regularly to provision Raspverry Pi-s.
You can use the standard unix 'dd' tool to copy images to sdcards and usb-drives. You can literally `dd if=whatever.iso of=/dev/rdisk3` (or whatever your disk id is). I have done it dozens of times. It doesn't give you a progress indicator, but I don't need to do this often anyway.
The pro version does make sense thought, I just can't imagine a long running manufacturing process where the SDcards are arriving faster than I can write to them.
Takes three clicks, doesn't require me to remember how to use dd, and prevents any stupid mistakes.
also what I like about etcher in particular is that it always seems to correctly identify usb drives even if some other tool has messed up the partitioning. I'm not the kind of person who can remember all the cli commands.
Unless you are a very rare person indeed, you presumably liked that your linux computer's OS had an installer to handle all the busywork for you, because installing the OS is not a project that interests you. Personally I am a big fan of compilers that prevent me from having to mess around in machine code. All of computing is abstracting and making things easier and I do not really get why Etcher in particular always attracts these 'i dont get it' comments when this exact microsm is the basis of our entire profession.
Because if you are creating some bootable media, it's not difficult to assume you are installing something on a raspberry pi, creating some bootable disk to mess around with your or someone else computer or phone, working on an embedded device, etc, etc. Basically I assume you know what a terminal is.
Maybe I am just too old, but I can't imagine downloading and installing a 430MB (just checked) GUI tool that runs one command.
Your example (gamers) makes sense though, and this should be a lesson for me.
A few minutes later...
"The dd utility copies the standard input to the standard output. Input data is read and written in 512-byte blocks. If input reads are short, input from multiple reads are aggregated to form the output block. When finished, dd displays the number of complete and partial input and output blocks and truncated input records to the standard error output."
Excuse me, what?
After you start dd, open another terminal and enter either:
`sudo kill -USR1 $(pgrep ^dd$)`
Or, if you're on BSD or OS X:
`sudo kill -INFO $(pgrep ^dd$)`
https://askubuntu.com/a/215521
but etcher seems to be way faster than dd for me
bs=1M will help you.
It’s not even about can you keep up with the rate manually it’s about is it cost effective, $1000 is what 50 hours at $20 per hour? That thing pays for itself in any setting.
Question: how do I justify it to myself? (kidding)
If you use Etcher as a way of switching the same installation drive to install a different OS by re-burning the install image, then Ventoy totally makes sense.
This looks awesome if I were to be making large rpi clusters
https://www.balena.io/privacy-policy/
https://forums.balena.io/t/serious-privacy-concerns-with-etc... (2018, also referencing a now-fixed software bug, but still an interesting look at what types of requests it makes by default)
- "We collect information about how you use our Service, such as the types of content that you view or engage with, the features you use, the actions you take, and the time, frequency, and duration of your activities"
- "Includes name of the device, operating system, and browser you are using"
- "Internet Protocol address, browser type and settings, the date and time of your request, and how you interacted with the Site"
Where in the code did you see this? I can't find that in their source.
I see they show a "banner" with this url: https://www.balena.io/etcher/success-banner?borderTop=false&...
show some featured apps: 'https://assets.balena.io/etcher-featured/index.html'
I can't remember if I used gnome multi disk writer or Balena etcher, but the "10x card readers on a 10 port hub" did the trick, and was relatively painless. there's no way my company was going to pay for this. Performance was limited by the SD card rather than the equipment.
Does each unit in the Daisy chain have a screen? Can you buy secondary units that don't have a screen? Seems to waste screens after the first one.
Also I'm really tired of growing acceptance of e-waste. "Burn 1000s of USBs at trade shows and throw them around" - QR codes are technology. Just print out a QR code that goes to an interactive website.
My biggest cost savings was in buying the cards on pallets, rather than having to shuck them from packaging individually. The time to stick all the cards in the slots, and pull them out afterwards, dominated all costs associated with duplication. If I had to do it again, and IF I could have frozen the image in time, I'd have just paid a duplication company to burn them. The cost for that was smaller than the cost of the cards.
And how do they know that? Correct. Telemetry.
I find the intersection of users who monitor network requests like a hawk but download 150mb apps to run dd hilarious.
I'm not even anti electron, you can pry vscode from my cold dead hands, but I've always been confused at the popularity of balena etcher in the first place.
and since it's open source, you can actually remove any home-calling and have a perfectly functional software, but you can't change anything on this website.
Etcher has and will always be an Electron app.
I can't guarantee that
That said, the fact that Etcher doesn’t require you to know this is a plus.
For comparison, the worst kind of Electron app in my opinion would be one that you have to keep open at all times, losing all the resources that Electron eats, on a "permanent" way.
-
Software is the core focus of our company balena (https://www.balena.io/). Formerly known as resin.io, we developed balenaCloud, a platform to build and manage fleets of IoT devices -- essentially removing friction when dealing with remote fleets (over-the-air updates, multicontainer application, etc).
As balenaCloud gained traction, one onboarding hurdle kept cropping up: people were struggling to flash OS images onto their chosen hardware platform.
Say you wanted to use a Raspberry Pi. You'd first need to use your terminal or some clunky software to download the desired OS image and flash it onto a micro SD card. Every OS had different software, forcing us to keep refining instructions, and none of the tools validated whether the write was done correctly. Next, you'd have to plug the SD card back into the Raspberry Pi and boot the system.
It might sound simple but the process was so tedious, slow and error-prone that it discouraged a good number of our potential users.
Frustration Necessity is the mother of invention and so we took the bull by the horns and made our own (open source) image writing application, Etcher, focused on user experience and speed. What we didn't know was that the problem extended far beyond our use case. As a result, Etcher now has a large community of users, supporters and advocates as it flashes about two million images per month.
-
We now had a slick image writer, but what happens when you scale your fleet? Etcher can flash multiple cards simultaneously, but the process still requires a custom USB hub with SD card adapters, which severely impacts your flashing speed. As for the available industrial solutions, the consistent feedback from our users is that these are expensive, slow and clunky -- everything Etcher is not. Further, none of these legacy duplicators support Etcher's ability to flash devices without an SD card.
The writing was on the wall: it became clear to us that if we wanted to help our customers scale their IoT fleets, we needed to build a hardware version of Etcher for professionals -- EtcherPro.
-
If this wall of text wasn’t enough, you can read the full blog post: https://www.balena.io/blog/taming-the-hard-in-hardware-in-8-...
or, you can watch an in-depth video presentation: https://youtu.be/pld8RSXu7ms?t=1178