Kaminario pushes its Cloud Data Platform

Kaminario, historical leader in high performance AFA, has confirmed its switch to a real multi-cloud strategy and presence. We had the privilege to organize a session recently with Kaminario to learn more about this adventure. The IT Press Tour in Israel was impacted by the Covid-19 and we did it via Zoom as usual now. We met 4 times Kaminario globally with the tour and this meeting helps us to measure all the progress made by the company for a few years now.

As the session started, we immediately felt a different atmosphere and dynamism when Dani Golan, CEO, jumped into it, refreshed us on the company and the strategy direction towards the cloud.

Kaminario realizes that the public cloud is an obvious choice for users with already 70% of organizations using cloud platforms today. Public cloud is clearly the preferred model and multi-cloud is mandatory to have some opportunities of success confirming the desire for users to be infrastructure agnostic able to offer the fastest time-to-market method to drive digital transformation.

The team has made a clear evolution towards a Cloud Data Platform and the product really rocks. Kaminario is a true believer of the composable infrastructure and promotes the decoupling of the storage from the compute layer to offer business agility and infrastructure flexibility. And to reach the desire state in 2025, Kaminario already started its journey with several key software developments with VisionOS, Clarity, Flex, NVMeoF, container and Kubernetes.

The first key element in the Kaminario approach is the notion of Data Pod which is VisionOS coupled with storage resources such devices and associated networks. Vision OS operates as the data hypervisor able to run on any cloud delivering enterprise class data services. Among them we find:

• DataShrink for data reduction with compression and deduplication with a high ratio,
• DataProtect with RAID, snapshot, replication and even AES-256 encryption, up to 8,000 snapshots are supported being just pointers,
• DataManage with GUI, CLI and RESTful API,
• DataConnect as an API to integrate vSphere, LogInsight, Microsoft VSS, OpenStack (it's dead), Kubernetes of course and Cisco UCS Director. Hum, why Cisco here, does it mean anything?

The second component is Flex that builds and manages Data Pods, it also assigns them to applications containers in private or public clouds and is able to move them across these clouds. Flex support s AWS, GCP and Azure in a BYOL model as Kaminario is not present at all on any of these cloud marketplace. And the last companion Clarity as the AIOps console to manage and optimize all environments just created.

The team, very early on, understood the need to make its software hardware agnostic to support any infrastructure especially with the cloud coming and most recently with containers and Kubernetes. These 3 key elements - cloud, containers and Kubernetes - coupled with NVMe over Fabrics transform a vision to a reality and makes Kaminario an storage software player fully agnostic of the environment. With such evolution, Kaminario represents today a leading SDS flavor probably one the most advanced on the market.

This approach is supported by a real compelling TCO, reducing costs by 30% over public cloud (example below compares AWS costs), leveraging a right provisioning mode and no cost for snapshot as the technique uses only pointers, meaning that the associated storage space is negligible.

What a journey since the AFA period with physical on-premises deployment a few years ago, the product was really good, to today with a pure software and cloud flavor. We'll monitor carefully Kaminario for the of 2020 as this model should have a real positive impact on their business and make them super attractive.