Collecting backups

Started by hainvv, Dec 01, 2022, 09:12 AM

Previous topic - Next topic

hainvvTopic starter

Tell me some solution. There are several hosting services, quite diverse. There is a home machine Ubuntu . All remote machines have SSH. Task: initiate an incremental backup of files and databases from a home machine and upload archives to it. It is desirable to store all data (logs, exception tables, etc.) also only on the home page.

Basically, the solutions found require the installation of an additional. software on remote sites, but this is not always possible, because there are shared hosting services. Therefore, most likely the maximum that can be used is tar and gzip.


If windows, install cygwin locally.
A primitive script of the form is written on the local machine:
scp user1@server1:/path1/files /* /local/path1/
scp user2@server2:/path2/files /* /local/path2/
And authorization by the key file is configured on the servers

Instead of scp, it is better to use rsync — more efficient in terms of traffic, especially when not everything changes in the file list (but to increase efficiency, rsync must be installed on servers as a daemon, this is not a problem for *nix)

maybe this service will suit you: .
In the free version, it gives 2 gigs of space for backups, collects them automatically according to a schedule via ftp or can connect to the database directly.
I use it myself for a couple of days, it seems to work fine


I "had the pleasure" here to backup 1.7 tb of photos in a 3-level structure using Borg — due to the fact that it stores a backup in binary files in a single array, listing could then take hours, and restoring hundreds of files for a couple of days is the same as the backup itself was originally performed. Worse, the feedback itself on errors or missing files was just as slow.

As I understand it, this is the problem of all solutions who store backups in the form of snapshots.
How is your utility doing with such a case? :)
Is it very difficult to verify the existence and restore several files from an incremental archive?