WeRecoverData

Tintri VMstore Data Recovery

Tintri VMstore is a hybrid or all-flash array built specifically around per-VM storage management rather than traditional LUNs or file shares, presenting an NFS datastore to hypervisors while tracking every virtual disk individually inside its own operating system. Tintri filed for bankruptcy in 2018 and its assets were acquired by DDN, which continues the VMstore line under the EC-series alongside legacy T-series hardware still in the field.

Platform Lineage and Naming

Tintri introduced VMstore with a hybrid flash-and-disk design, later extending the line to the T400, T500, T600 and T800 series and the higher-capacity T5000 and T7000 all-flash systems. Tintri OS underpinned all of these with per-VM analytics and management as its distinguishing feature.

Following Tintri's 2018 bankruptcy, DDN acquired the company's assets and continued VMstore development under its own branding, releasing the EC6000 and EC8000 series while maintaining compatibility with the per-VM model pioneered by the original product.

Generations and Models We Evaluate

Generation / familyModels
Legacy TintriVMstore T400, T500, T600, T800, T5000, T7000
DDN-eraTintri EC6000, EC8000

Architecture and Data Layout

VMstore abstracts storage at the level of individual virtual machines and virtual disks rather than LUNs or volumes, so administrators and Tintri OS itself track per-VM performance, snapshots and capacity even though the underlying datastore is presented over NFS.

FlashFirst places active data on flash and de-stages colder blocks to disk on hybrid models, with Tintri OS making that placement decision continuously based on observed per-VM I/O patterns rather than fixed tiers.

Snapshots and clones in VMstore are taken at VM or vDisk granularity, which is useful for restoring individual machines but means that recovering a whole datastore after serious damage requires reconstructing the per-VM metadata layer, not just the NFS export.

Protocols and formats: NFS, VMware vSphere, Hyper-V (select models), REST API for per-VM management

Failure Scenarios

Logical failures

Hardware failures

What Not To Do Before an Evaluation

Our Evaluation and Recovery Process

Frequently Asked Questions

Is Tintri VMstore still supported now that DDN owns it?

DDN continues the VMstore line under the Tintri and EC-series branding; legacy T-series units in the field may have limited vendor support depending on age and contract status, which is worth confirming before planning a recovery.

Can individual VMs be recovered without touching the whole datastore?

Because VMstore tracks storage at per-VM granularity, isolated VM or vDisk recovery is often possible where the underlying array and its metadata are otherwise intact.

Does the per-VM design change how RAID recovery works?

The physical RAID layer is recovered in a broadly similar way to other block arrays, but an additional step is needed afterwards to reconstruct the per-VM mapping so individual virtual disks can be identified.

Related Platforms and Services