Redesign Disks section to group datasets under a shared pool header
Closes #682 (closed)
Summary
- Replaces the misleading display where each dataset card showed identical pool-level storage numbers (size, used, free, progress bar) with a grouped layout
- One pool header shows the true shared ZFS storage stats (total size, progress bar, used size + percentage, free percentage)
- Below the header, each dataset gets its own card showing dataset-specific info: mode, clones, snapshots, size
- When a pool has no child datasets (e.g.
dblab_pool_01), the redundant dataset name is suppressed — the pool header already identifies it
Edited by Denis Morozov