Project Status¶
MeshStor CSI is in Technical Preview. The data path works end-to-end and has been exercised against the failure modes documented in Self-Healing and Volume Relocation. The CRD schema and StorageClass parameters may change in backwards-incompatible ways before 1.0.
Do not yet store data here that you cannot afford to recreate.
MeshStor has not yet published its own benchmark numbers. Any third-party numbers cited in the docs — for example in the Comparison table — come from published sources and are attributed inline. For the architectural overhead analysis, see Performance.
Stable and unstable surfaces¶
| Surface | Status | Notes |
|---|---|---|
| CSI provisioning interface | Stable | It's CSI |
| StorageClass parameters | May change | Names may be reorganized before 1.0 |
MeshStorVolume CRD schema |
May change | Field names and structure may shift |
MeshStorNodeDevice CRD schema |
May change | Field names and structure may shift |
| Reconciler internal state machine | Experimental | Internal — operators should not depend on specific phase names |
| Deployment manifests | Stable | StatefulSet + DaemonSet pattern is settled |
What's coming¶
On the open-source roadmap¶
- Snapshots and clones — planned, but expect a real cost at snapshot creation. The current MD RAID + GPT data path doesn't natively support copy-on-write; the planned implementation temporarily adds one extra RAID member, takes a sub-second
xfs_freeze, then removes the temporary member. After the snapshot exists, subsequent writes have no per-snapshot overhead — this is the inverse of LVM thin /dm-snapshot, which pays the snapshot tax on every write. - Raw block volumes — planned. The current driver only supports XFS-formatted volumes; raw block mode requires a small change to the CSI volume capability handling.
- Ext4 filesystem — planned. Currently XFS only.
- Additional Linux distribution support — Ubuntu 26.04 LTS and Rocky Linux / AlmaLinux 10.x are planned. See Compatibility for the currently tested matrix.
Available to paid customers¶
These features are not on the open-source roadmap. They are reserved for paid customers.
- Encryption at rest
- Volume expansion when
drivesPerCopy ≥ 2(RAID10 expansion). Online expansion is supported today only whendrivesPerCopy=1. - KubeVirt live migration support — implemented as a controlled multi-attach shim that lets two pods on different nodes hold the same volume during the live-migration window.
- Enterprise Linux distribution support — Red Hat Enterprise Linux, OpenShift (RHCOS), SUSE Linux Enterprise Server, SLE Micro, and Talos Linux. See Compatibility.
Not planned¶
- True ReadWriteMany. MeshStor is a block storage driver. RWX is not on the roadmap and not under consideration. If your workload needs real multi-writer access, use a file storage system instead.
How to engage¶
- Source code: github.com/meshstor
- Issues and discussion: file an issue against the repository.
- Support expectations: Technical Preview is best-effort. There is no SLA for the open-source build.
What's Next¶
- Overview — the three differentiators that define MeshStor
- Limitations — the canonical list of unsupported features and roadmap status
- Comparison — head-to-head against alternatives
- Use Cases — when MeshStor is and isn't the right tool