OpenZFS: Why the Filesystem Matters
ZFS is a copy-on-write filesystem that checksums everything. Every read verifies data integrity. Silent corruption — the kind that eats backups and isn't discovered until restoration fails — is detected and corrected automatically when you have redundant vdevs.
Key properties:
- Atomic snapshots with zero performance overhead at creation
- Send/receive for efficient replication to remote hosts
- Compression built-in (lz4, zstd) — often faster than no compression because reads are smaller
- Datasets with individual quotas, inherited properties, and mount options
- RAID-Z (1/2/3): analogous to RAID 5/6 but without the write hole
ext4, XFS, and btrfs are good filesystems. ZFS is in a different category for environments where data integrity and replication strategy are first-class concerns.
pf, pfSense, and OPNsense: The Network Stack
pf (packet filter) originated in OpenBSD and spread to FreeBSD and the BSDs generally. pfSense and OPNsense are FreeBSD-based firewall/router distributions built on pf. They power a significant fraction of small business and homelab networks worldwide.
The pf rule syntax is clear and auditable. Stateful filtering, NAT, traffic shaping, and CARP for high-availability failover are all well-supported. If you're running a pfSense or OPNsense appliance, you're already in the BSD ecosystem — you might not know it.
bhyve and Jails: Virtualisation Without the Overhead
The bhyve hypervisor is built directly into the FreeBSD kernel: mature, lightweight, and well-integrated with Jails for a hybrid container/VM model that predates modern Linux container ecosystems by years. Jails provide strong OS-level isolation without the complexity of a full container runtime.