Home > Storage > PowerScale (Isilon) > Product Documentation > Storage (general) > PowerScale OneFS Operating System > Overview
The most important design choice and fundamental difference of Dell PowerScale scale-out NAS is that with OneFS the storage system does not rely on hardware as a critical part of the storage architecture. Rather, OneFS combines the three functions of traditional storage architectures—file system, volume manager, and data protection—into one unified software layer. This combination creates a single, intelligent file system that spans all nodes within a storage system.
The Dell PowerScale storage nodes provide the appliance hardware base on which OneFS runs.
While the hardware consists of industry-standard, enterprise-quality components produced by manufacturers, such as Intel, Seagate, and Mellanox, nearly all distinctive aspects of the storage system are provided in OneFS software. On this commodity hardware base, OneFS enables data protection and automated data balancing and migration, and the ability to seamlessly add storage and performance capabilities without system downtime.
Dell PowerScale OneFS clusters can be architected with a wide variety of node styles and capacities, in order to meet the needs of a varied dataset and wide spectrum of workloads. These node styles encompass several hardware generations and fall loosely into four main categories or tiers. The following table illustrates these tiers, and the associated hardware generations and models:
OneFS works exclusively with the Dell PowerScale and Isilon storage system, referred to as a “cluster.” A single Gen6 cluster consists of one or more chassis, each containing multiple storage nodes. The nodes are constructed as rack-mountable enterprise appliances containing memory, CPU, networking, 40 Gb Ethernet or QDR InfiniBand, and storage media. The cluster can scale out as high as 252 nodes.
In addition to the modular architecture, such as the PowerScale H700, where four nodes reside in a 4RU chassis, OneFS also supports the stand-alone all-flash PowerScale F910 NVMe, F900 NVMe, F710 NVMe, F600 NVMe, F210 NVMe, and PowerScale F200 nodes. Both the chassis-based platforms and the stand-alone nodes happily co-exist within the same cluster.
The OneFS total single file system capacity easily spans from 10’s of terabytes to 10’s of petabytes, supporting individual files up to 16 TB in size. Each node added to a cluster increases aggregate disk, cache, CPU, and network capacity. As a result of this aggregate increase, a 252-node cluster can access as much as 181 TB of globally coherent, shared cache. The cluster can also support up to 186 PB of raw storage capacity—before any data-reduction savings. With capacity and performance delivered in a single storage system, a single file system, and a single volume, the complexity of the system and management time for the storage administrator does not increase as the system scales.
OneFS stripes data across all the storage nodes in a cluster. As data is sent from client machines to the cluster (using industry-standard protocols, such as NFS, SMB, S3, HTTP, and HDFS), OneFS automatically divides the content and allocates it to different storage nodes in parallel. This occurs on the private Ethernet or InfiniBand network, which eliminates unnecessary network traffic. The cluster is managed as a single file system and the coordination and data distribution are completely transparent to end-user clients. When a client wants to read a file, OneFS will retrieve the appropriate blocks from multiple storage nodes in parallel, automatically recombining the file, and the initiating client sees exactly what was originally written. This ability to automatically distribute data across multiple nodes in a transparent manner is fundamental for the ability of OneFS to enable growth, next-generation data protection, and extreme performance.