zfs backup question

Julien Cigar <[email protected]> Fri, 19 Jun 2026 10:58:58 +0200

Newsgroups	gmane.os.freebsd.questions
Message-ID	<ajUC39Hhy3JGOEnK@x1>

Hello,

We are backuping our servers with Sanoid / Syncoid. 
Our servers provde access to several (web) applications and each 
of them is stored in a dedicated ZFS dataset and running in a 
dedicated jail. 
So we have something like zroot/jails/app1, zroot/jails/app2, 
zroot/jail/appx, etc

Due to some changes to our CI/CD pipeline and the way our apps are 
deployed, the zfs dataset (zroot/jails/xxx) is now destroyed and 
re-created on each deployment (the "app" is distributed as a gzipped 
zfs dataset and the deploy process is a simple send/recv). 
It works well except for backuping as there is no common ancestor 
anymore each time the dataset is destroyed.
The answer is probably "no", but: is there a way to handle this 
without having to send the whole dataset when backuping?

(We are thinking of maybe use Podman (or maybe simply a FreeBSD 
package?) to package/distribute the application but I haven't found 
any good documentation on how packaging something with Podman yet ..)

Thanks,
Julien

-- 
Julien Cigar
Belgian Biodiversity Platform (http://www.biodiversity.be)
PGP fingerprint: EEF9 F697 4B68 D275 7B11  6A25 B2BB 3710 A204 23C0
No trees were killed in the creation of this message.
However, many electrons were terribly inconvenienced.

signature.asc (application/pgp-signature, 833 B)