A backup rotation filter for the Unix shell(git.sr.ht) |
A backup rotation filter for the Unix shell(git.sr.ht) |
May I suggest changing the name to something more direct? Calling it "--invert" means the user must think through what the default sense of the test is, then negate that in their mind. Not exactly a tough mental task, but people do make careless errors.
Perhaps something like "--list=keep" and "--list=discard" (with the default being "discard").
Also, typos:
"will make prunef to keep" --> "will make prunef keep"
"list backups that will should be" --> "list backups that should be"
I will consider changing the invert flag, but I'm not that happy with something like "--list=...". There will be only two modes with discard being the default one. So imho there should be only one flag to switch to the non-default mode.
Meta: I'm excited to see sr.ht starting to pop up in the wild like this:) I hope this is part of it starting to take off.
[1] https://borgbackup.readthedocs.io/en/stable/usage/prune.html
"... for the Unix shell" makes little sense. When I used Unix my shell was csh, later tcsh. Nowadays my shell is bash in most cases and dash in some more limited environments. Either case, "the Unix shell" does not exist.
tar -zcv <SRC_BACKUP> | gpg -c --batch --passphrase <PASSWORD> -o <DEST_BACKUP>.gz.gpg
* /var/backup/db/db1/monday.sql
* /var/backup/db/db1/tuesday.sql
* /var/backup/db/db1/monday.sql
For databases that change more frequently I instead backup every 1, 3, 4 hours as appropriate:
* /var/backup/db/db2/monday/00.sql
* /var/backup/db/db2/monday/04.sql
The appeal of this is that I always have "local" backups, and I don't need to consider rotation at all each one gets the most recent copy when it runs, and I have an alert/alarm to make sure files are recent enough that things aren't broken. I appreciate that if your databases dumps are 600Gb each, or something similarly sized you'd waste a lot of space, but for small things the simplicity of this approach is a good win.
(These get copied offsite as part of the backup of the whole filesystem. In the past I used to backup only some stuff, that failed the first time I tried to restore a mailserver and didn't have /var/lib/mailman archived! These days I explicitly backup "/" excluding only /tmp, /proc, /sys, and /dev.)
(Zfs snaps make awesome backups too, but unlike tar are inherently tied to zfs)
If you sign the backups with some distinct key of the backup server, why wouldn't the attacker have access to those keys too (in the above scenario they already have access to the keys that the backup server is using for encryption).
I know there's an open issue in age for adding authentication[1], so there clearly is some threat this would protect against but I can't figure it out.
It's fine though, I leave most blame at iOS' door. Everything looks too big on this temporary iPhone SE, and I'm not allowed to zoom out, default to Firefox, or use any extension or 'content blocker' in it even though it's forced to use Safari to render. (/Rant..)
I think this is just some missing CSS to hand this case.
> Everything looks too big on this temporary iPhone SE, and I'm not allowed to zoom out, default to Firefox, or use any extension or 'content blocker' in it even though it's forced to use Safari to render.
On my very much not temporary iPhone SE, I can use content blockers and zoom out…
And I mean 'zoom' out from the default, e.g. on desktop I browse most sites at 80% in FF, some 67 or 50, fewer at 100.
On my in-for-repair Android phone, the smallest system UI/font setting is smaller, and FF is allowed extensions and to use its own renderer, so I have control over that.
Everything just seems like I'm using a largified accessibility mode. And typing - impossible to place the cursor mid-word? So if suggested corrections are wrong, no choice but to delete and re-type the whole thing. And no select all? So if I decide not to post such a rant, I have to fumble with the two cursors and move one to each end myself.