Reinstall OS


Still waiting for this feature …


This be a great feature i agree. :blush: because pay for every time when you destroy/create new server is little to much. :scream:


Is there an ETA on this feature?


i see that request is not realy importat,sad.:neutral_face:


+1 for this so common feature… :frowning:


@edouard @jcastets Could you comment on this? I’d also like to know if this could be added in future updates and what the ETA might be.


Hello @dodedodo,

The re-install OS feature is not planned at the moment. The process to re-install the OS at the moment is the following:

  • Terminate the previous server
  • Start a new server with the Image you want
  • Remap the public IP previously owned by the server to the new one

Note: All public IPs are reserved IPs by default.



Hello @edouard,

That only sounds good on paper but doesn’t work well in practice. I have just deleted a c1 to reinstall the OS. Now your system is telling me c1 is out of stock.

What we need is a feature where we have more stock.


Yep, I can reinstall OS more than five times a week while training setting up something, and paying for every server re-creating isn’t what I expected when choosing Scaleway

+1 without a doubt :cherries:


I cannot believe scaleway is missing such a basic but important and HUGE issue.
I thought with scaleway my dream comes true. But no, i have to use providers that allow for this simple button.

You are making your customers jump through totally unnecessary hoops.

This missing button is making my team crazy and we will abandon scaleway if this is not fixed within a short timeframe.

I apologize for hurting anyones feelings but sometimes the truth is hard.

Even as a money making scheme this doesn’t make any sense at all.

Same goes for a button “Disable SSH Key login and allow root password login” by the way.
It makes testing automated scripts a lot easier and spares the user from constantly fixing sshd_config.
But yes, that’s not the topic here :wink:

My apology for ranting hard…but really guys, fix it! And possibly allow for selecting multiple instances and reinstall them all at once. I would even be ready pay a few dimes for that.BUT PLEASE, make this possible.



I don’t think they care


Sure they do. But it’s not a big issue. If you have such strong needs to reinstall the OS, you have bigger problems than a missing button.

The greater problem with the OS reinstall at scaleway is more to do with the disappearing server, where releasing a server for reinstall can potentially leave you without new stock to reinstall on. Of course this can be remedied by grabbing a second server before releasing the original. But new comers to scaleway may get tripped by that.


The availability problems are a whole different subject. Right now you can’t rely on scaleway to scale (no pun intended…) to your needs whenever you want.

Some of us try out new software, and it’s very valuable to be able to start over with a clean install. I agree that on production systems it isn’t that much of an issue, but that’s simply not the only use case.

Also, if online did care they would have given us at least a timeframe in which this feature would be implemented. It’s been 11 months since the start of this thread. So no, they don’t care.



I’m just starting to eval Scaleway for a small development team who will be creating and destroying systems a lot. The issue of additional charges depending on at what time of the month a system is rebuilt means that Scaleway has an issue that should be fixed.


As this is an simple option but scaleway and is taking it as an big task, Before the Digital Ocean have the same problem but they have implemented an button to install the os by the way the users can install the OS when ever they needed. I think Scaleway or created this community for reduce the ticket created for this type of features. Because they don’t care about the customer’s and their requirements. If they continue like this definitely everyone will leave from their service. If Any of the responsible person from Scaleway and is looking this post/thread. act now other wise Iam dam sure we will leave from your serve. See may be you are selling the service at low cost but think every one is not looking about the cost but also the support and features .

So awake Scaleway or tell your mama ceo or responsible person to implement it soon :grin:


Why is it so hard to just make a big button ‘Reinstal’’ -> Choose OS -> it takes care of it.

DigitalOcean has the same functionality, so why is this suddenly rocket science?


Need this feature too. Please add it !!!


Don’t be like profits over service. Add it please. Or is it added? :confused:


I noticed the annoying procedure (and the slight upcharge) around a year ago, but didn’t really bother grumbling. This isn’t a convenient place really, it’s a place that’s cheap, and thus has to be mostly automated so that it can run on it’s own. It does feel like it’d be more annoying than it price, but hey, I don’t own or operate this place.

But now it’s different. Stock is constantly low or out. The procedure no longer works - after “Step 1. Delete the server”, you are now stuck without a server. You can sit around for arbitrary lengths of time hoping to snag a second instance so that you can install and then release the first, but that becomes ever rarer as more and more people actually need what they have. It’s a first-come first-serve system and that’s ok, but reinstalling or switching OS shouldn’t shove you to the back of the line.

This makes it extremely sketchy to set up anything you plan on having here for a while. If it requires any fix deep enough to require reimagining, you can’t really count on it. If you just want to work with images rather than fiddle inside individual systems (as is becoming the norm and we all know we probably should) that’s also a no-go. I already have a vultr instance on the side. I still haven’t totally thrown in the towel on armv8s and might even be able to containerize things into a single larger instance down the line but… I gotta scale, and even if option #2 was now rather than future, it won’t magically conjure in working order - it’d need reimaging.

Just sayin’. Make of it what you will.

  1. Back up any critical data.
  2. In web ui, under server settings, show advanced and change boot script to rescue.
  3. Reboot server
  4. Use normal scaleway ssh key to log in via putty - web console won’t work
  5. Reformat your volume, ie mkfs.ext4 /dev/vda
  6. Mount your volume to mnt, ie mount /dev/vda /mnt
  7. Copy rescue base to volume, ie rsync -ahPHAXx --exclude={/dev/,/proc/,/sys/,/tmp/,/run/,/mnt/,/media/*,/lost+found} / /mnt
  8. Back in the web gui, change your boot script back to a normal one.
  9. Reboot your server - watch it in the web console.

Server should start as expected.

Worked for me with Xenial on a VC1S.