backup-now script overwrites previous instances when multiple DBs configured

[androidfans]

Description

In create_schedulers() at install/assets/functions/10-db-backup, the fix for the >> append bug (which caused duplicate backup entries on container restart) replaced >> with > for writing /usr/bin/backup-now:

echo "/usr/bin/backup${instance}-now now" > /usr/bin/backup-now

This means when multiple DB instances are configured (e.g. DB01, DB02), each loop iteration overwrites the previous content. Only the last instance ends up in backup-now, so backup-now only triggers the last configured database backup.

The same issue exists for the individual backup${instance}-now scripts a few lines above:

cat <<EOF > /usr/bin/backup"${instance}"-now

Expected behavior

Running backup-now should trigger backups for all configured DB instances (DB01, DB02, etc.).

Suggested fix

Initialize the file with > once before the loop (or truncate it), then use >> inside the loop:

# Before the loop
> /usr/bin/backup-now

# Inside the loop
echo "/usr/bin/backup${instance}-now now" >> /usr/bin/backup-now

This avoids both the original append-on-restart bug and the current overwrite bug.

Version

4.1.100 (also present in 4.1.99, carried over from tiredofit/docker-db-backup)