The Nutanix Virtual Computing Platform delivers all the benefits of traditional arrays, with exceptionally low latencies for both random and sequential workloads.
This high-performance converged solution combines storage and compute resources into a single, integrated system. The next-generation system eliminates the need for an external storage network, such as a SAN or NAS. Data I/O is local to the server, yet behaves like centralized storage.
Each 2U Nutanix appliance has four high-performance server nodes, integrating best-in-class Intel CPUs, high-speed memory, and local storage (SSD and HDD). Each node runs a Nutanix controller VM, forming a fully distributed cluster with the same capabilities of centralized storage.
The Nutanix Virtual Computing Platform is a scalable virtualization platform for desktop, server, and big-data deployments. Every node comes pre-installed with a hypervisor and is ready to run guest VMs out of the box.
IT teams can quickly provision VMs with VMware vCenter, or any other virtualization management solution such as Microsoft System Center Virtual Machine Manager (SCVMM).
The Nutanix platform is VMware-ready, and supports all virtualization capabilities, including live VM migration (vMotion), high availability (HA), dynamic resource scheduling (DRS), and fault-tolerance (FT). The Nutanix solution is fully certified to run popular virtual desktop solutions, including VMware View and Citrix XenDesktop, as well as traditional server and big-data workloads.
Storage resources from individual server nodes in the Nutanix cluster are virtualized and converged to form the Nutanix Distributed File System (NDFS). This storage platform has a single namespace and is accessible by all hosts, appearing as centralized storage.
Administrators can simply provision VMs to the Nutanix datastore, and NDFS takes care of the rest.
NDFS simplifies storage configuration. There is no need to manually define storage volumes, LUNs, or RAID groups, as traditional storage arrays require. Once the appropriate storage containers and datastores are initially defined for the environment, the virtualization manager does not have to provision the appropriate storage resources for each new VM.
Nutanix Information Lifecycle Management (ILM) handles data placement and moves data locally to where I/O originates. This takes the network out of the I/O path and ensures the highest possible performance.
ILM combines with Nutanix heat-optimized tiering (HOT) and automatically moves
frequently accessed data up to the highest performance storage tier (PCIe SSD).
When data becomes idle or “cold,” it is moved down the storage tiers to
maximize storage efficiency.
The Nutanix platform continuously analyzes data-access patterns and intelligently
moves data to the best storage tier for the fastest performance overall.
The Nutanix Virtual Computing Platform integrates enterprise-class management capabilities to handle nearly any virtualization environment. Whether you are deploying a VDI solution, setting up a new private cloud, or provisioning a new workload for test and development, Nutanix makes it simple.
The Nutanix platform provides native VM-centric snapshots for converged backup, as well as array-based quick-clones for efficient provisioning.
The platform supports full integration of the snapshot and cloning capabilities of VMware by supporting both the VMware API for Array Integration (VAAI) primitives and View Composer for Array Integration (VCAI) standards. This integration increases the speed of operations by offloading them to the Nutanix platform.
The Nutanix Virtual Computing Platform provides two forms of compression to allow flexibility and optimize capacity by workload.
For sequential workloads: The system compresses data synchronously as it is written using in-line compression. That process optimizes capacity and eliminates the performance impact of the sequential I/O.
For random or batch workloads: Data is written uncompressed. As it becomes “cold,” data is moved to the lower-performance storage tiers where post-process compression is performed. This provides the best performance by not impacting the normal I/O path and compressing only when data and compute resources are available.
The Nutanix platform allows customers to manage compression policies in a truly VM-centric fashion. Unlike traditional solutions that compress data mainly at the LUN level, the Nutanix solution provides all of these capabilities at the VM and file level. It is more
efficient and simpler than traditional architectures.
You get the highest possible compression/decompression
performance with sub-block-level granularity.
Along with compression, Nutanix offers native thin provisioning,
which optimizes capacity use by allocating physical storage resources
only when they are needed. Nutanix rounds out these benefits
by offering efficient snapshots and QuickClones, a more efficient
native clone of VMs, which aids in restoring VMs from backup.
Nutanix offers native, VM-centric snapshots and replication for disaster recovery (DR) and backup. Administrators can configure replication policies that specify groups of VMs in a primary site. These can be replicated to any combination of secondary sites to ensure maximum service resiliency and application performance.
In addition, Nutanix provides a storage replication adapter (SRA) to support VMware’s Site Recovery Manager (SRM). Customers, who are accustomed to SRM technology for their virtual environments, can seamlessly utilize the SRM graphical user interface to manage VM-centric disaster recovery from Nutanix.
Any Nutanix cluster can serve as both a primary and secondary site simultaneously for different VMs, which means flexibility and highly efficient use of resources.
Data protection and availability
The Nutanix architecture protects both data and metadata. Its evolved RAID-like protection stores all blocks of data across multiple nodes and disks.
This system ensures that no data is lost in the case of a failed controller, disk, or node.
The distributed Nutanix architecture provides a virtual storage controller per node. Metadata is distributed among all controllers in the cluster, thus eliminating any controller as a single point of failure.
Nutanix supports auto-pathing, which makes the failure of any virtual controller completely transparent to all hosts by allocating another controller to assume responsibility for an affected host. The result is a platform with zero service interruption.
The Nutanix architecture incorporates the same design principles and technologies as the Google File System (GFS), which powers Google’s massive web properties and applications. This design pools local compute and storage resources into a tightly integrated system, and ensures that all VMs can access available storage anywhere in the cluster, from any node.
New nodes are dynamically added to the cluster, scaling compute and storage capacity. The n-controller model eliminates the performance bottleneck of legacy storage architectures that have a finite number of controllers. It ensures predictable and consistent performance as the environment scales without the need to re-architect the infrastructure.
The Nutanix Virtual Computing Platform auto-discovers new nodes, providing seamless cluster expansion. This functionality uses a Bonjour-based protocol for auto-detection and integration of new nodes. Administrators are able to quickly and easily expand a cluster without any downtime.
The Nutanix Prism management framework delivers a single-pane graphical interface for both compute and storage. All information is organized and presented via elegant touch points to facilitate easy consumption of operational data. Built from the ground up using HTML5, Prism provides the ability to define and manage a complete converged infrastructure from nearly any device. Additionally, Prism includes a programmatic REST-based API for integration with third party cloud management systems.
The user interface enables cluster expansion with just a couple clicks and provides an at-a-glance dashboard to monitor performance and latency on a per-VM basis. Dynamic alerts notify admins of hardware failures and capacity issues.
These events and alerts are mapped over time and overlapped with performance and usage-analysis data. Datacenter managers see intelligent summaries in real time tiles so they can monitor the health and status of the cluster.