just starting with a new OH3 installation, and my idea was (as before) to use the good, old “dd” command to make fully backups (remove the SSD to another linux machine and dump the OH partition into a file). This was working in last three years fine, including restores!
Now I see Amanda, but I don’t understand how this tool should be able to make a fully working full-backup including Grafana and a running influxdb.
I already found this docu: openhabian/openhabian-amanda.md at main · openhab/openhabian · GitHub
But for me still unclear
My target is to save the OH-Partition to a another storage including ALL things (like a real snapshot (1:1 copy) from a stopped system).
In my understanding this is not pssible with Amanada, becasue the OH (and DB) are running, or where is my fault ?
Since you’re asking, your fault is to misinterpret what Amanda is.
It is not a replication tool like dd but a backup system. It in fact is possible to use it also to create a one-time copy of a raw partition but if that is all you want (you didn’t explain why you would to do this) then dd is a simpler hence better choice.
If your intent is to have a system that allows to restore your box at any time to more or less any state throughout time (which is what backup systems do and what is usually meant when someone refers to a “backup”) then dd is a bad choice and you should go with Amanda.
Thx Markus for the answer,
my dd (raw backup) will be the backup for worst case.
While installing new bindings, testing rules and Dashboard etc., I can do a openhab-cli backup, maybe daily.
What is the different between the cli-backup (started manually when i think is usefull/needed) and the options from Amanada?
Can Amanda store more data, configs ?
openhab-cli only stores openHAB configuration.
Amanda does all of the system, including OS and 3rd party software you might run.
That’s not a backup worth speaking of as one. You need to keep it up to date to be useful.