TL;DR: Have a FreeBSD 15+ system? Run pkg install sylve and follow the
getting-started guide to bring up your first node.
Sylve is an open-source infrastructure management platform with a simple goal: compute, storage, networking, and day-to-day operations should work as one system. It brings Bhyve virtual machines, Jails, ZFS storage, networking, backups, and cluster operations into one web interface. It uses what is already built into FreeBSD instead of layering a separate appliance on top.
Sylve v0.3.0 is now available. With more than 500 commits since v0.2.3, it is our largest release so far. The biggest change is how much better the different parts of Sylve now work together.
We are very happy with how the product is maturing. Sylve is being used in real-world enterprise environments, and that feedback has pushed us to spend a lot of time on the things people notice when something goes wrong: recovery, validation, observability, safer changes, and predictable failure handling.
Where Sylve is today
v0.3.0 lets Sylve manage much more of a host without sending you to a collection of separate tools.
Networking saw some of the biggest changes. Sylve can now manage PF traffic and NAT rules, reusable network objects, routes, and WireGuard. Firewall logs, counters, service-aware telemetry, and generated configuration previews make it easier to see what the host is actually doing.
VMs and Jails can now be migrated between nodes in the same cluster. Sylve transfers the bulk of the ZFS data while the workload is still online, then performs a final incremental transfer and guarded cutover. There are preflight checks, progress reporting, recovery state, and guards against conflicting storage or lifecycle operations.
We also put a lot of work into storage and day-to-day operations. Backups now commit related datasets and VM disks as coordinated generations, with stronger restore and recovery workflows. The release adds iSCSI initiator and target management, deeper ZFS and ARC telemetry, scheduled S.M.A.R.T. tests, and notifications through the interface or ntfy, email, and Discord. A protected local CLI and interactive console provide another way to carry out common VM, Jail, networking, and lifecycle work.
Certificate and Dynamic DNS management are built in now, too. Each node that needs one can use a free Sylve.app hostname and managed TLS certificate without opening an inbound validation port or sending its private key off-node. Cloudflare and Namecheap are also supported as Dynamic DNS providers, with more integrations planned.
Replication is available as an opt-in early-access feature in the v0.3.x series. It is backed by extensive automated testing across failover, fencing, recovery, and cluster-state handling, and has also been exercised in real-world enterprise environments. We call it early access because we are rolling out a failure-sensitive feature carefully, not because it is an untested prototype. We still recommend using it alongside independent backups. The full v0.3.0 release notes cover the complete set of changes, compatibility notes, and upgrade details.
Why try it
You do not need to be a FreeBSD specialist to use Sylve. It is built around familiar infrastructure tasks: running workloads, managing and sharing storage, connecting services, monitoring systems, and protecting data. FreeBSD supplies the building blocks underneath: Bhyve, Jails, ZFS, PF, and a capable base system. Sylve puts a shared model and interface around them, so creating a workload, attaching storage and networking, monitoring it, backing it up, and moving it to another node do not have to be separate projects.
It also stays small. The Go backend and web interface ship as one service, with a baseline memory footprint below 384 MB. Sylve itself can run on FreeBSD 15 without packages outside the base system.
A few extra packages are used only when their features are enabled: libvirt for VMs, QEMU tools for disk images, Samba for SMB shares, swtpm for TPM emulation, and dnsmasq for DHCP and DNS. The FreeBSD package and port install these for you. A Jail-only or otherwise narrower setup does not need to run the whole stack.
You can try Sylve on modest hardware and still grow from one host into a clustered setup later. It is not a thin dashboard over unrelated services. The same system understands how workloads, ZFS data, networking, backups, and cluster ownership relate to each other.
Try it in your browser
You do not need a FreeBSD host to explore the interface. The demo on sylve.io is a working three-node cluster with sample data that runs directly in the browser. Use the embedded version on the home page or open it full screen.
Support what comes next
The FreeBSD Foundation supported work across both the previous release and v0.3.0 through its project grant. The grant paid for visible features, but also for the integration, testing, and reliability work needed to make those features useful together. We report progress through the FreeBSD quarterly status reports.
The Foundation has since expanded that support in two related directions. It awarded Sylve another grant for the next phase of the platform, and it is sponsoring memory ballooning support for Bhyve. Developers at Alchemilla are actively working on both. One stream builds out the management experience, while the other improves the underlying virtualization stack for Sylve and the wider Bhyve ecosystem.
For Sylve users, the new grant will bring native OCI container management through Podman, an App Gallery with repeatable installers for Jails and containers, and faster deployment using application templates.
The same effort expands Sylve as a storage platform, with NFS sharing, automatic ZFS dataset tuning, mobile monitoring, operating-system updates from the interface, and ZFS scrub schedules. It also includes end-to-end testing and clearer migration documentation.
We are also designing a simpler interface for people moving from TrueNAS CORE 13.x and its FreeBSD-based storage model. We do not see Sylve as only a VM platform or simply a counterpart to Proxmox. Virtualization is important, but future releases will also make Sylve a complete NAS and infrastructure management system.
Beyond the grant-funded work, Alchemilla handles ongoing development and maintenance. IPTechnics contributes infrastructure and real-world enterprise testing. For v0.3.0, Senex IT also made a one-time contribution through GitHub Sponsors that helped fund part of the release.
Different sponsors bring different things to a project like Sylve. Storage, virtualization, networking, and cluster operations do not look the same in a home lab, at a service provider, or inside an enterprise. No single organization sees every use case. Grants, long-term development support, infrastructure, and operational testing all help in different ways. Together, they give us a better view of what people actually need.
That support also produces work that is useful outside Sylve. One example is gzfs, an open-source project that grew out of the same work. gzfs provides a typed Go interface for ZFS, ZPool, and ZDB operations, including JSON parsing and test-friendly command execution. It is useful on FreeBSD, but it also supports other OpenZFS-compatible systems and can be used entirely independently of Sylve. Supporting Sylve can therefore produce tools that help the wider ZFS and open-source infrastructure ecosystem too.
If this approach to infrastructure management is useful to you or your organization, there are several ways to take part. You can test v0.3.0 in your environment, share operational feedback, contribute code or documentation, provide testing infrastructure, or sponsor continued development. Organizations interested in supporting a particular area can also contact us. A wider mix of community and sponsor support gives maintainers the time and stability to improve Sylve and its supporting projects over the long term.