Firstly: my apologies if this is in documentation or forums somewhere, but I’ve not been able to find it after a while of looking.
There’s the option when configuring Virtualmin > Backup and Restore > Scheduled … [the backup itself] to have the destination as “Amazon S3”. What I can’t work out is if / how to use an Amazon S3 compatible service that is not actually Amazon S3 - for example, Object Storage - Get Started | Linode
Is it possible to set up with Vitualmin?
Cheers!
Matt
PS: I did find System Settings → Virtualmin Configuration → Backup and Restore but all I can spot there that might be related is “Full path to aws command for S3 operations” and no info on what that is or means.
rclone is a command-line utility that will backup stuff to, or retrieve it from, almost any well-known cloud storage destination. It takes anywhere from less than a minute to a few minutes to configure, depending on the destination. Once configured, you can append it to a backup routine or run it as a cron job.
I think @Joe is contemplating wrapping a GUI around rclone. I think it would be a good idea for those who don’t like working in the shell because it’s so versatile. But configuring it from the shell is actually very easy if you’re comfortable in there.
As for S3, I’ve never used it on Virtualmin, so I really can’t comment on the implementation. In general, you basically have to create a bucket in S3 first, and then copy the required data and paste it into the transport application. I think you need the bucket name, API key, and endpoint; but I haven’t set up an S3 bucket in years, so I could be wrong.
Note that “S3 compatible” is a fuzzy term. Not everybody really is compatible enough for the library we use to be able to work (but, I think quite a few do work at this point).
Agreed, however I find “most” of the popular providers implementing an S3-like service are taking Amazon’s lead so “generally speaking” should work properly.
Well depending what do you looking for an strategy you can implement…on my server I run local backup on same server I have disk just for that, then I use off-site backup, just Linux OS no Apache nothing just pure os and terminal as extra package to os base, no gui at all. Now I know it’s not aws but costs per tb are very low.
Basically bash script running each night from vmin production server copy with rsync everything outs to be backed up to remote location (SSH keys connected only, so no need change ports no need for fail2ban etc) once command finished, just another start to send me telegram notification that remote backup was done and then it will kick local backup on HDD which is outside of vmin but still mount as another local HDD, like usb… I keep rolling backups for server in order 5days rolling back on local disk mout (disk space 4tb) and remote backup daily for about month (13tb disk space)
Regards security, local HDD have no protection at all of course but remote backup have script which allows delete command once unknown ip connect from outside of my lan, basically it will kick SSH key deletion before it will let you load terminal and type any command in. Of course you have to have on remote server separate SSH key for your admin or your self. Once compromised, you can take server offline, investigated and on remote server you can fix issues and then just pull backup back to your newly reinstalled production server.
This set-up will cost you cents and you are free setup your strategy as you want anywhere even your home pc or laptop… aws is great don’t get me wrong but linode or my firends server, who cares, just deploy anywhere plus take costs of the setup is well above 10 dollars a month on aws for just the HDD space…think about it. Little bit of time and coffee would do
Edit: I just suggest this as alternative and all know how is out there but if you Sysadmin my suggestion would make a sense somehow. Thank you.
Thinking of this because it’s the best cost option in our instance, as opposed to expanding the disk space available on the VPS. Linode’s deal is a good one.
