"DO WHAT I MEAN!!!" syndrome, plain and simple.Autocompletion should magically understand your custom script takes a zfs dataset as argument?
"DO WHAT I MEAN!!!" syndrome, plain and simple.Autocompletion should magically understand your custom script takes a zfs dataset as argument?
zfs snap -r $(dirname $(df -h /usr/local | awk 'NR == 2 {print $1}'))@old
According my (limited) experience, I prefer an integrated solution like ZFS respect a composition of different storage layers components, like in Stratis.I haven't heard of Stratis before and it looks similar to ZFS, but I'm curious about the difference with Stratis formatting the new pool entry with XFS (standard Linux FS), vs ZFS seemingly being the filesystem and pool management all-in-one? Could it loosely be seen like ZFS and Stratis being similar, but ZFS being a lighter option without needing a separate filesystem format?
This last paragraph sounds a lot like the old divide between RAID cards and the filesystems that sat on top of them, i.e. one layer not knowing a single thing about the other, and all the limitations that arose from that.Moreover, ZFS code can take care in an integrated way of every aspect of storage management: e.g. caching, checksum, different types of disks, etc.. On the contrary, every component of Stratis is "blind" about the low level details managed by other components, and so some optimizations and integrity checks are more difficult (or impossible) to be implemented in Stratis.