--keep-exclude-tags and other borg flags? #234
Labels
No Label
bug
data loss
design finalized
good first issue
new feature area
question / support
security
waiting for response
No Milestone
No Assignees
2 Participants
Notifications
Due Date
No due date set.
Dependencies
No dependencies set.
Reference: borgmatic-collective/borgmatic#234
Loading…
Reference in New Issue
No description provided.
Delete Branch "%!s(<nil>)"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
What I'm trying to do and why
I'm trying to convert my silly shell scripts into nice little borgmatic configs.
Steps to reproduce (if a bug)
I have this shell script:
I could find most flags, but not
--keep-exclude-tags
. I can't figure out how to pass arbitrary flags down to borg.Other notes / implementation ideas
Maybe there could be an "extra flags" setting of some sort to just shove stuff down into borg?
Environment
borgmatic version: 1.2.11-1 (note that borgmatic --version gives an error)
borgmatic installation method: debian package
Borg version: 1.1.9
Python version: N/A
operating system and version: debian stable 10 (buster)
I implemented support for
--keep-exclude-tags
via a new borgmaticlocation
section option calledkeep_exclude_tags
. You should be able to upgrade to borgmatic 1.4.4 and set the option totrue
. Thanks for suggesting it!As for a way to pass arbitrary flags to Borg.. I've instead been opting to add support to borgmatic for new or missing flags on-demand. That way borgmatic ultimately gets native support for everything Borg can do. Although I realize that can be somewhat obnoxious in the interim.
Please feel free to file tickets for any other missing options, and we can always revisit this approach if need be!
That's probably because
borgmatic --version
was added in 1.2.14. Time to upgrade! 😄it's great that borgmatic supports so many of borg's option, but this is bound to create needless complexity and synchronization issue.
it also makes it needlessly hard to keep both programs in sync: for example now I would need to uprgade borgmatic (to a version not yet packaged in debian, let alone backports) to benefit from this (great!) new functionality. if there would be support for arbitrary flags, i wouldn't need to upgrade at all!
inversely, if I want to try out a new borg feature without support from borgmatic, it's not possbile directly from borgmatic unless I get it implemented there first...
i would encourage you to reconsider, but i understand the reasons why you're going this way. :)
it would be nice to have a comparative table to list which features are supported and which aren't, if you keep going that way...
thanks for the extremely fast response!
I'd be willing to entertain support for arbitrary flags if you file a ticket for it, especially if this is something several users are interested in.. I just haven't seen a super strong need so far. But then again, I'm the one in the position to add borgmatic support for Borg flags as soon as I need them. 😃 I realize that doesn't apply to everyone!
There actually is a script included with borgmatic which tries to probe for flags that are not yet supported, but it's written in hacked-together shell and therefore doesn't work super reliably.
As for the topic of upgrading: I unfortunately don't have the resources to support all old versions of borgmatic. For instance, 1.2.11 is nearly a year old, and there have been 39 releases since then, some of them quite significant! I totally understand why you might be using Debian stable, and the version of borgmatic included with it.. but I encourage you to find a way to upgrade! Example: Switching to installing directly with pip.
If there are things I can do on my end to make using newer versions easier (keeping in mind limited resources), please let me know!
there are ways to run newer versions of stuff in debian with having to revert back to pip and so on. the normal way to do so is through "backports", but first the newer release needs to trickle down into unstable, testing and then stable-backports. since not even 1.4.0 has hit unstable yet, this might take a while. :)
but i'm not in a rush.
thanks!