Flash is indeed very effective at addressing random IOHwever disk tends to be better at delivering sequential IO on a per dollar basis.Also the price differential between flash and disk is between 10-20x (depending on type of flash and type of disk).Flash as an endpoint of storage is still unproven (data integrity, bit error rates etc) whereas disk has been around a lot longer and is much more matureCallout at the bottom is what this slide leads uptoie an ideal solution marries the strengths of flash and disk (what we refer to as hybrid).

The improvements you gain are marginal at best.If otoh, you can architect a file system from a clean slate, you have the ability to leverage the best of flash and disk and deliver the performance, capacity, and intelligent data management delivering a well-rounded solution.

Part 3 of the positioning:All storage vendors offer flash optimized storage however not all flash optimized storage is created equal.A majority of the ones offered by incumbents layers on flash over an architecture optimized for spinning disk.

Bottomline: if you can architect a system which allows you to flexibly vary performance, you can cater to the needs of all mainstream workloads with a single storage architecture Specifically the performance demands of the workloads vary – eg VDI/OLTP much more performance centric vsSharepoint.

Cost economics is not a major factor here.Long term archival and retention will likely move more to the cloud moving forward.The middle chunk is best addressed by hybrid flash optimized storage. These are workloads that demand us latencies and >1M IOPS. Bulk of the mainstream workloads are in that middle chunk.We believe the top tier will move to flash resident on server or all flash arrays. Point to note is that both the top and bottom tier of workloads are nichey in nature.

Next part of the positioning gets into workloads and how these were addressed in the pre-flash era.

THIS IS AS PER GARTNER’S APPROVED TEMPLATE Make sure to check with HQ on the latest #s.The idea is to highlight how Nimble is experiencing growth all around and is being lead by storage industry veterans with a strong track record of success.

Introduce NimbleThe headcount, # of customers and deployments, as well as the # of VARs changes frequently.

Introduce yourself and the topic of discussion.Network 4 x 1GbE 6 x 1GbE or 2 x 10GbE + 2 x 1GbE (G models) IO Modules Dual hot-swappable 2x 6Gb SAS-2 controllersġ: Effective capacity is usable capacity after no to 2x (or 50%) compression typical range: 30% to 75%Ģ: The CS420 model is available in CS420-X2 and CS420-X4 options only.Ĭontrollers CS210 CS220/G CS240/G CS260/G CS420/G(2) CS440/G CS460/GĬontrollers High Performance (15K– 20K IOps) Extreme Performance (35K – 60K IOps)Īdd on shelves Add 1 shelf Add up to 3 additional shelvesĬapacity(1) 4-9 TB 8 – 16TB 17 - 33TB 25 - 50TB 8 – 16TB 17 - 33TB 25 - 50TB Nimble Storage CS-Series and ES1-Series ProductsĬapacity(1) 11 – 22 TB 23 – 45 TB 34 – 68 TB –Lucas Clara, IT Director, Foster Pepper LLC A mailbox or Exchange system can be recovered in literally minutes.īest of all, we can regularly test our procedures for Disaster Recovery.‖ Require very little space and are recoverable and mountable locally and Exchange users notice no performanceĭegradation. ―With Nimble, we were able to run 3 snapshot backups a day and –Ron Kanter, IT Director, Berkeley Research Group

Very quickly we realized we hadĮnough performance headroom to consolidate our very demandingĮxchange 2010 deployment on the same array.” ―We started out deploying SQL workloads primarily on

MAILBOXES PER DISK Published and verified – Jared Call, Director of IT, Imagine Learning With other disaster recovery products, it’s veryĮffective and very competitive in terms of Schedule and meet our data retention policy.‖ Our application or storage performance at all.‖Ĭouple clicks of the mouse to establish a backup 17#NimbleStorageįast and non-disruptive – they don’t impact

Nimble san disks