Open-E JovianDSS Proxmox Plugin 1.0.0
We are pleased to announce version 1.0.0 of the Open-E JovianDSS Proxmox Plugin.
Version 1.0.0 marks the first stable release of the plugin and the transition from the 0.x development series.
This release is based on 0.11.6 and introduces no additional functional changes. For the complete list of changes included in the underlying release, see the 0.11.6 release notes.
Plugin Overview
The Open-E JovianDSS Proxmox Plugin integrates Proxmox VE with Open-E JovianDSS storage and allows administrators to manage storage-backed VM and container disks directly through standard Proxmox interfaces.
The plugin supports:
- QEMU/KVM virtual machine disks and LXC container volumes
- Dedicated JovianDSS volumes for individual virtual disks
- Thin provisioning
- Volume creation, deletion, resizing, and movement between guests
- Space-efficient snapshots
- Snapshot rollback
- Full and linked cloning
- iSCSI connectivity with optional multipathing
- NFS-backed storage
- Proxmox cluster operation
- Sharing a JovianDSS Pool between multiple Proxmox clusters when unique target prefixes are configured
- Cluster-wide coordination of concurrent storage operations
- Automatic recovery from selected iSCSI export and connectivity failures
- Proxmox VE High Availability environments
Getting Started
New users should begin with the following guides:
- Quick Start — prepare JovianDSS, install the plugin, and configure the first storage.
- Plugin Installation and Updating — installation, upgrades, and removal.
- Plugin Configuration — available storage settings and configuration options.
- Networking — network requirements and recommended topology.
- Multipathing — resilient iSCSI connectivity using multiple paths.
For additional operational and recovery guides, see the complete JovianDSS Proxmox Wiki.
Known Issues
Orphaned target after changing target_prefix — Deleting a volume after changing the target prefix may leave an empty orphaned target on the JovianDSS appliance. This is a cleanup issue only and does not affect data or connected devices.
Clone cleanup after VM restore — After restoring a VM from a clone, the clone may remain attached to the iSCSI target. This is related to Proxmox bug #6879. It should not affect normal operation because the clone is removed automatically during a subsequent shutdown or migration.
NFS snapshot cleanup — Snapshots of NFS-backed volumes may require manual deletion after the associated VM is removed.
Cancelled snapshot operations — When a snapshot operation is cancelled, Proxmox VE may terminate plugin cleanup before it completes. This can leave the VM locked, with a partial snapshot in the prepare state and snapshot-related resources present in JovianDSS but not registered in Proxmox VE.
To recover, run:
qm unlock <vmid>
qm delsnapshot <vmid> <snapname> --forceThis removes the remaining snapshot resources, including the vmstate disk.
Disk import, export, and cross-cluster offline migration — These operations are currently unavailable. A volume-name parsing issue will be addressed in a future release. In addition, the Proxmox ZFS migration workflow relies on ZFS send/receive, which is not supported by the current JovianDSS plugin. Use Proxmox Storage Migration to move volumes between storage systems.
Notes
Unique target prefixes
When sharing a JovianDSS Pool between multiple clusters — or, more generally, between multiple Proxmox storage definitions — each storage definition must use a unique target_prefix.
The plugin enforces this requirement: it refuses to share an iSCSI target between storage definitions and identifies the storage definition that must be reconfigured.
Proxmox High Availability
Virtual machines and containers managed by Proxmox High Availability are governed by HA policies. Direct state-changing operations, such as rollback or migration, may conflict with active HA management and trigger recovery actions such as automatic restart or failover.
Before performing these operations, temporarily disable HA management for the affected VM or container, or set its HA state to ignored. This prevents HA from interfering with the requested operation.
Failures caused by active HA restrictions are expected Proxmox behavior and do not indicate a storage-plugin malfunction.