This is the research completed at the digital gardening bee 30th June 2018—it is yet to be implemented:# <blockquote>1. Use mysqldump to make a backup </blockquote><blockquote>https://www.mediawiki.org/wiki/Manual:Backing_up_a_wiki#Mysqldump_from_the_command_line# 2. You’ll want to save to a different filename every time, e.g. </blockquote><blockquote>nice -n 19 mysqldump -u $USER --password=$PASSWORD $DATABASE -c | nice -n 19 gzip -9 > ~/backup/wiki-$DATABASE-$(date '+%Y%m%d').sql.gz </blockquote><blockquote>You can skip the “nice -n 19” parts of this command, but the useful part is the date.# </blockquote><blockquote>3. Test that the backup works </blockquote><blockquote>https://www.mediawiki.org/wiki/Manual:Restoring_a_wiki_from_backup </blockquote><blockquote>I’m not super clear on how to test.# </blockquote><blockquote>4. Passwords are hashed and salted, so it’s probably ok. Use scp/sftp to copy, e.g. </blockquote><blockquote>scp wikifile.gz remote.server:~/backups/ </blockquote><blockquote>5. You can put all that in a batch script and then use crontab. See </blockquote><blockquote>https://www.mediawiki.org/wiki/User:Flominator/Backup_MW</blockquote>