As I write this, Tintri (NASDAQ: TNTR) is trading at .28 a share and falling. It’s speculated that the storage company will declare bankruptcy with no possible acquisition in sight.
Tintri’s claim to fame is how easy it can be integrated into VMware environments and allowing for VM level storage management. While novel when it first arrived on the scene, since vSphere 6 and the introduction of Virtual Volumes, the company lost what made it unique.
Although it could be possible another company can swoop in and acquire the company for the intellectual property they have at the last moment. Then again, it may be worth having a serious conversation on alternatives if the inevitable happens and if you’re a customer, you’re left with a major piece of infrastructure with no support.
Pure announced that it will be replacing it’s entire M series line of all-flash storage arrays with the X series NVMe arrays. Prior to the announcement, the M series FlashArray used SATA based flash modules while the only NVMe based array, the X70 used the company’s new NVMe drive modules.
Now, the M series arrays will no longer be sold and taking their place will be the X10, X20, and X50 arrays. A new X70 will also be sold and a new flagship, the X90 has been intoduced.
Currently Pure does not support end-to-end NVMe as there isn’t support yet for NVMe-OF interfaces. However, there are plans to support it on these new arrays. Additionally, the lower tier arrays, the X10 and X20 support SATA flash but are DirectFlash ready, which is Pure’s name for NVMe flash. This way customers can still take advantage of an all-flash array without having to do a forklift upgrade if they need to upgrade to NVMe storage.
Current customers can take advantage of the new arrays through their evergreen storage program.
With Pure, NetApp, DellEMC, and HPE all announcing arrays that have or will support NVMe on flash storage, only Pure is supporting NVMe on all arrays rather than just on the high end. Pure looks to make the technology more mainstream rather than an exclusive feature.
That wasn’t the only thing announced. Pure is working with nVIDIA by using their FlashBlade scale-out NAS with nVIDIA DGX-1 servers for artificial intelligence development. Announced was the introduction of the AIRI Mini. AIRI stands for “AI Ready Infrastructure”. The AIRI mini is a system with 7 blades capable of 364 TB of data effective, 700k NFS IOPS, and has 16 GPUs that can perform 2 PFLOPS.
If you asked me five years ago what I thought about hyperconverged systems, particularly Nutanix, I probably would have dismissed it. Most of the players then were niche, startup companies all vying for the attention of anyone willing to give them a chance.
After a couple years, I realized how wrong I was as larger companies were taking notice. For instance, VMware introducing vSAN. Later on, Dell partners with Nutanix to sell their software on Dell servers.
This year I had the opportunity to attend Nutanix .NEXT, their annual conference which was held right down the interstate from me in New Orleans.
First, I have been very impressed on how quickly Nutanix has grown to be a major player in datacenter infrastructure in addition to their software offerings that continue to disrupt how we do information technology. It’s not just about replacing traditional infrastructure with scalable nodes in a software-defined stack. It’s about getting you to shift the way you do IT from the traditional client/server model to a web scale cloud infrastructure that obfuscates the notion of on-premises and public cloud. Basically, what’s underneath a presentation layer is practically invisible.
A few of the announcements during the conference I thought were worth noting:
Nutanix Beam – Beam is a Software-as-a-service, the first of its kind for the company, which allows organizations to manage multiple clouds, including on-premises cloud on Nutanix infrastructure. Additionally you can do cost analysis, regulatory compliance, and security compliance from a single console. The big feature that really impressed me wasn’t just showing my periodical spending with public cloud but how much I could save if I’m just wasting resources and where I can cut such waste. Even better, with Nutanix’ Xi cloud services you can take advantage of AHV or vSphere deployed in Google Cloud Platform.
Nutanix Flow – Nutanix’ answer to VMware’s NSX, only easier to manage and implement. While it may not have the feature parity of NSX, it does include the capability to do microsegmentation that allow organizations to further secure applications. Everything is easily managed from Prism Central.
Object Storage Services – Nutanix AOS has been webscale for quite a long time, not surprising that object storage is now supported. If you have a need for S3 compatible storage, have Nutanix nodes in your datacenter, and don’t want to manage yet another storage system your prayers have been answered. OSS also includes support for encryption, role based administration, quotas, and WORM.
Further improvements to AHV – Acropolis Hypervisor is the plucky new kid on the block that has been looking more and more like an enterprise heavyweight to finally break the duopoly of ESXi and Hyper-V. In the 5.5 release announced earlier this week you should see vGPU support, RBAC, and AHV turbo mode. The latter is an optimized data path between the VM and storage. More Nutanix customers are starting to adopt the KVM based hypervisor as they find it supports many of the same features they need from the other two, without the associated cost.
Other than that there were plenty of excellent sessions and courses with plenty of information. Additionally, other partners of Nutanix were on hand to show how their products integrate into Nutanix’ offerings. I certainly enjoyed the conference and hope to have the opportunity to attend next year. If anything, Nutanix has shown they will not let up despite their lead in the hyperconverged infrastructure market.