Tintri upgrades all-flash arrays, adds scale-out functionality

T-series arrays acquire 3D flash drives, with the T5060, T5080 and T5040 all set to benefit

Tintri has announced a refresh of its T-series all-flash arrays based on upgraded 3D Nand flash drives, plus scale-out capability for its all-flash and hybrid flash arrays.

The hardware upgrade has been applied to the T5060 and T5080 all-flash arrays, as well as the T5040 entry-level flash box.

The 2U arrays can take flash drives of 0.5TB, 1TB, 2TB and 4TB. Depending on which drives are fitted, capacity ranges from 6T to 23TB in the T5040, 12 to 46TB in the T5060 and between 25 to 92TB in the T5080. That’s raw capacity, which Tintri claims is boosted by up to 5:1 data deduplication ratio.

Input/output (I/O) performance ranges from 100,000 input/output operations per second (IOPS) to a possible 200,000 IOPS. Quality of service (QoS) operates at a per-virtual machine (VM) level, as does replication.

Tintri arrays connect as file access via a network file system (NFS) or server message block (SMB) and offer storage in native format to virtualisation hypervisors, so doing away with traditional storage provisioning formats such as the logical unit number (LUN).

The new scale-out functionality is part of an upgrade to the Tintri Global Center management suite, which can manage up to 32 Tintri nodes. It will allow customers to define a pool of storage of up to 10 Tintri nodes that can be treated as a single pool for optimisation purposes.

Using scale-out functionality, Tintri claims customers can scale up to around 10PB of capacity and host storage for up to 160,000 VMs.

Scale-out functionality allows multiple physical storage devices to be connected and for data to be retained and managed as a pool, with policies and actions applied across them.

“We optimise in an intelligent way, so we won’t tie up the storage for hours migrating data or losing snapshots or policies associated with VMs”
Chuck Dubuque, Tintri

It arose initially in the network-attached storage (NAS) space, where scale-out systems offered the ability to link many NAS boxes together where previously discrete systems had become siloed, with data inaccessible.

Tintri’s scale-out functionality is based on the Global Center’s ability to track stats at virtual machine level from all nodes in the pool, such as whether, for example, sets of drives are thin provisioned.

On the basis of this, the Global Center (a storage control platform) can make least-cost recommendations to migrate data from one set of media to another to relieve I/O hotspots.

Managing storage

“The key advantage is to enable higher utilisation by using intelligent VM-level storage optimisation,” said Chuck Dubuque, product marketing director at Tintri. “But we make sure to optimise in an intelligent way, so we won’t tie up the storage for hours migrating data or losing snapshots or policies associated with VMs.”

Tintri supports VMware, Microsoft Hyper-V, Red Hat Enterprise Virtualization, and Citrix Xen virtualisation platforms on its arrays and can support them all in a scale-out pool. However, at this time scale-out will only make recommendations for migration to optimise between VMware and Hyper-V.

Will this eventually extend to optimisation and migration between all supported hypervisors?

“It’s possible and it’s something we have demoed internally,” said Dubuque. “But it’s not on our roadmap yet. We’ll be looking to input from our customers on how valuable that is.”

Read more about all-flash and hybrid flash

Read more on Virtualisation software