Currently I'm running backups (on Linux) through rsync and backing up the entire "~/.config/joplin-desktop/" folder.
But, as I am really paranoic with my data, and as I am also starting to use joplin in a "heavy" way, I would like to know your opinions, please, about this method: is this correct?
I.e., supposing I need, in the future, to restore data, all I would have to do would be to "install" the Joplin's Appimage and to put the "joplin-desktop" folder in its correct place? Doing that I would have all my data up and running again?
That should work and is a quick way of restoring that Joplin client.
I would also suggest that you additionally look at the Simple Backup plugin. This backs up everything to a folder of your choice, with a frequency of your choice, keeping as many versions as you choose and can encrypt those backups (useful if the backup location is a Cloud Sync folder and the backup will be transferred off-site).
Simple Backup saves all your notes to a Joplin Export File (JEX). A JEX file is in fact just a TAR file, however the contents are all your notes AFTER they have been exported from the database. So as well as your resources it has all your notes in MD format, NOT in a database. A JEX backup comes in handy if things have gone really wrong and you have to wipe everything, including the sync target, and start again. I say that as if you import a JEX backup into a Joplin client that still has those notes it will duplicate them not replace them.
I have previously realised that some time back I had deleted a note that I probably shouldn't have. I just opened an old JEX file created by Simple Backup and searched the MD files for text I knew was in the note. Then I just copied that Markdown and pasted it into a new note.
Thank you very much, @dpoulton . This Simple Backup plugin you've mentioned seems to be really great.
It's very good to khow your experiences while using the plugin. I'll definitely put this plugin in my backup routine, and that's something I really wanted: the ability to have all my notes also in the format they are after exporting from the database.
I'll now run my backups both via rsync and Simple Backup plugin.
No. I believe that the Joplin export uses the unique IDs so that it can maintain the links, tags, parent notebook etc. when imported back into Joplin. The title of the note becomes the first line of the note MD file.
I'm on Windows, but would like to add that as a relatively new user, one time I did a version update, only to discover Joplin was then only showing the new version files, and all my own files were gone.
Fortunately through Settings > Sync > Advanced, I was able to Delete local data and re-download from sync target
From then on I always backup with a .jex file, and all is good.
There are 2 approaches to backups and it depends on your setup.
client side
Simple export
It is easy to export your notes. While you don't want to do that daily, it is a good strategy on first installs to do a manual export after "big changes" and until you have a better solution. Keepo reading.
Plugin
Many mentioned the Simple Backup plugin above and this is definitely utltra simple to setup so this is IMO the first thing to put in place, no matter what is your plan and setup. This will at least protect your data for the time being. Ideally, your backup target is not the same disk/raid array than where you joplin client is installed. The best option is to use a NAS, an external hard drive, etc... or have some automation to fetch this local backup and move it away from the same disk.
server side
Here things become more complicated due to the number of options Joplin supports.
Cloud backup excl. joplin-server
If you use a cloud backend, you shoud ensure of proper backup of that. Some NAS like the Synology ones offer a cloud backup solution that fetches data from our cloud accounts and backs it up locally.
joplin-server
Here again, it depends...
It depens how your joplin-server is setup.
If ALL your data lands in a postgresql db, you need to backup your postgres database and there are plenty of guides usually based on a cronjob to create a dump and delete older backups.
if your data is split between a database and filesystem, I feel like stating the obvious but those will both need to be backed up and kept together as you will need the FS backup matching the DB backup.
Word of caution
Most of us know it but how many do it... A backup is ONLY a backup if you tested a restauration of it... That means wiping your data and restoring your last backup to see how it goes.
I ran into an issue I mentioned here while doing this so you may want to test the restore with a seond profile&account to avvoid "polluting" your main account.
