Features

Store any kind of unstructured data, on any hardware, at any scale, and for any workload

Features

Overview
Store
Protect
Process
Serve

Overview

OpenIO overcomes all the limitations of traditional object stores while offering better functionalities.
Store
- Scale-out by nature. A shared nothing model aggregates capacity from independent x86 and ARM servers.
- Limitless storage based on open source technology.
- Store 1,000+ PBs of data and hundreds of billions of objects.
Protect
- Data Replication with multiple copies, local and/or remote.
- Reed-Solomon Erasure Coding.
- Various topologies from a single data center to multiple or stretched clusters across geos.
Process
- Install and run applications where data resides.
- Better application SLAs with data locality.
- Flexible application support with APIs, file-based access, or SDKs.
- Unparalleled TCO reduction.
Serve
- Native Object APIs: C, Python, and Java.
- Standard Object APIs: OpenIO REST/HTTP, Amazon S3 and OpenStack Swift.
- Multiple file sharing and access methods: NFS, SMB, FTP and local FUSE.

Key features
Data management
Data access
Grid for Apps
Platform management
Compare

The most flexible object storage on the market
- Innovative scale-out design
  
  A grid of nodes, instead of traditional cluster layouts, flexible, resource conscious and capable of supporting heterogeneous hardware.
  
  Nodes in a cluster can be different from each other, allowing different hardware generations and capacities to be combined over time without affecting performance and efficiency.
- Hardware agnostic
  
  A fully software-defined solution that runs on x86 and ARM with minimal requirements (starting at 1 CPU core, 512 MB of RAM, 1 NIC, and 4 GB of storage). Supported in mixed physical and virtual environments as well as deployed in containers, the cloud, or at the edge.
  
  Freedom of choice and cost savings for any type of deployment.
- Event-driven
  
  OpenIO SDS catches all events that occur in the cluster and can pass them up in the stack or directly to an application on the local node. OpenIO’s Grid for Apps serverless framework takes advantage of all OpenIO SDS features and can process all its events seamlessly, and third-party frameworks can be integrated as well.
  
  An application backend for micro services, automated data processing workflows, and event-driven applications.
- Multitenancy by design
  
  The cluster can have multiple domains as well as accounts and containers. Data can be organized in different storage pools and protection schemes.
  
  Segregate workloads and improve overall system efficiency.
- Dynamic placement policies
  
  Thanks to Conscience technology, each operation is performed on the most available node. Each node of the cluster computes a quality score every few seconds and shares it on a distributed shared board accessible to all nodes.
  
  Each time a new operation is requested, the best nodes are selected for optimal resource utilization, continuous system tuning, and enforcement of customers fine-grained constraints.
- No rebalancing
  
  Conscience technology eliminates the need for cluster rebalancing. When new nodes are added, they progressively offload nodes in place thanks to the scoring mechanism. Their quality score is better than for older, heavy-loaded nodes, and they’ll be used more often.
  
  New resources are immediately available ensuring that performance improves instantly.
Data management
- Data placement
  
  OpenIO SDS is designed to be efficient while simplifying data workflows and continuously optimizing available resources.
  
  Storage pools
  
  Groups of storage media, hard disks or SSDs, can be configured to isolate particular traffic or objects from others. Data can be moved from one pool to another with simple operations or through automated tasks, enabling automated tiering. This improves multitenancy, performance consistency, and eases the organization of data across different media pools.
  
  Dynamic Storage policies
  
  User-defined policies allow SDS to select the best storage pool and data protection scheme based on object size. It improves data footprint efficiency and overall system performance on mixed workload environments.
  
  Data compression
  
  Data can be compressed at rest, synchronously or asynchronously to select the best tradeoff among data footprint reduction, i/o ops optimization, and CPU consumption. The platform operator is able to choose the best fit for his use cases. Compression saves space and improves overall system efficiency.
  
  Automated tiering
  
  Data can be automatically moved across different storage pools thanks to an elaborate lifecycle management rules system (based on metadata, size, tags, prefixes, ect). This simplifies data movements in the system, leaving faster resources ready for new hot data.
  
  Hybrid Cloud
  
  Automated tiering can be expanded to the cloud. OpenIO SDS currently supports BackBlaze B2. Infrequently accessed data can be offloaded from the system to a cheaper, remote repository.
- Data security
  
  OpenIO SDS is designed to store data safely, making it secure when in transit or at rest.
  
  WORM
  
  Write Once Read Many can be enabled on containers, disabling the ability to modify or delete objects. It allows to keep data safe from deletion and comply to regulations that need this requirement.
  
  Object encryption
  
  All objects can be encrypted on the fly, before leaving the access layer for the internal network to improve intra-cluster communication.
  
  Encryption at rest
  
  All data stored in the system can be encrypted, improving overall system security.
  
  Updated
- Data life-cycle
  
  OpenIO SDS offers seamless and powerful data life-cycle management capabilities.
  
  Delayed Compression
  
  Time-based policies can be set to compress data that has not been read for some time, improving space efficiency and $/GB for cold data.
  
  Paid plan
  
  S3 life-cycle compliance
  
  OpenIO SDS supports S3 object life-cycle management, allowing end users to set up time-based rules and trigger actions accordingly.
  
  Automated life-cycle data management
  
  Automations can be set to move, compress, delete data by age, last modified date and many other characteristics, improving system efficiency and data placement for best performance and capacity utilization.
  
  Paid plan
- Data protection
  
  OpenIO SDS allows to choose between several protection schemes for all cluster layouts and user needs.
  
  Multiple object replicas and Erasure Coding
  
  Multiple object replicas and erasure coding are selectable by the end users for any account or container, and can be easily changed, providing a wide range of choice to protect data.
  
  Container snapshots
  
  Container snapshots enable to make quick copies of entire containers for backup, to create data sets for big data analytics or test applications. That speeds up backup operations and offloads large data copies to the storage backend.
  
  Data integrity checks
  
  Everytime an object is written in the system it is also read again to verify its integrity, and the same operation is repeated over time to ensure that data is safe. Avoids silent data corruption.
  
  Dynamic data protection
  
  By adopting dynamic storage policies the best data protection is automatically chosen by looking at the type or size of each single object. X3 replicas for small objects, short erasure coding for medium sized objects, large erasure coding for big objects. Highly improves efficiency.
  
  Distributed Erasure Coding
  
  The system can distribute erasure coded objects geographically for the best space efficiency across multiple data centers and good Disaster Recovery capabilities.
  
  Paid plan
  
  Geo replication
  
  Policy-based synchronous and asynchronous data replica are supported. Stretched clusters, leveraging distributed erasure coding alongside asynchronous replication for long distance, high latency and low bandwidth connections. Build highly resilient and distributed storage infrastructure that can spread around the globe.
  
  Paid plan
Data access
Strong S3 compatibility

Strong Amazon S3 compatibility to access data easily from any application, gateway, or third-party solution.

Updated

Openstack Swift API

Compatible with OpenStack Swift and applications using this protocol to access data. Allows SDS to replace Swift.

OpenIO SDKs

Native Object APIs for Python, C, and Java. Developers can use their programming language of choice to integrate their solutions with OpenIO SDS.

REST/HTTP API

Native open source API that can take advantage of all the characteristics of OpenIO SDS for applications that require advanced storage features and deep integration. Easy to use, it takes advantage of all SDS features.
OIO FS

File System connector that links OpenIO SDS to storage entities through NFS, SMB, FTP file protocols, and local FUSE mountpoints

Paid plan

Third-party email connectors

Allow email servers and collaborative software solutions to access SDS object stores directly: OpenIO SDS can be used as primary and/or secondary storage option for such platforms. Currently available:

Cyrus IMAP 3.0 connector

Zimbra connector
Grid for Apps

OpenIO SDS allows to process data at scale, directly on the storage infrastructure.

Metadata Indexing and Search

Real-time event-based metadata indexing and search. The power of ElasticSearch combined with the flexibility of OpenIO SDS.

Updated
Platform management

A comprehensive set of tools to deploy and manage OpenIO SDS clusters of all sizes with ease.

Deployment Tools

A complete set of pre-built scripts designed for easily installing and configuring large-scale clusters simplifies deployment and configuration activities in large scale deployments.

Paid plan

Command Line Interface

Easy to use and consistent CLI to manage the cluster and simplify system administration. The quickest way to administer an OpenIO SDS cluster.

WebUI

Graphical User Interface that offers an overview of cluster status and enables easy monitoring for day-to-day operations without using the CLI.

Paid plan Updated
Why Object Storage?
Block

File (NAS)

Object Storage
- Location
  Block: Data center, close to servers
  
  File (NAS): Local network, close to clients
  
  Object Storage: Internet, accessible from everywhere by any device
- Protocol
  Block: FC, iSCSI
  
  File (NAS): SMB, NFS
  
  Object Storage: HTTP, APIs
- Data/Access ratio
  Block: N:1 - 1:1
  
  File (NAS): 1:N
  
  Object Storage: 1:Millions
- Performance
  Block: High IOPS, Low latency, Consistency
  
  File (NAS): Good IOPS, High latency, Good throughput
  
  Object Storage: High parallelism, Higher latency, High throughput
- Scalability
  Block: Usually scale-up (less than 1 PB)
  
  File (NAS): Scale-up + limited scale-out (multi PBs)
  
  Object Storage: Scale-out (100+ PBs, Exabytes)

Start today
Grow as you need

Community Free
- Community support
- OpenIO SDS Open Source
- S3 and Swift Gateway
- OpenIO Command Line
- OpenIO SDKs
Get started
Standard $/Capacity/Year
- Standard support
- All Open Source features
- Geo replication
- Deployment Tools
- Web UI
See plan
Premium Standard plan
+ Fixed price/Year
- Premium support
- All Open Source features
- Geo replication
- Deployment Tools
- Web UI
See plan

What do you do with your data?

Experience OpenIO today, and discover how your business can be more agile and unlock the value of your unstructured data.

Get started

Features

Store any kind of unstructured data, on any hardware, at any scale, and for any workload

Features

The most flexible object storage on the market

Data management

Data placement

Data security

Data life-cycle

Data protection

Data access

Grid for Apps

Platform management

Why Object Storage?

Start today Grow as you need

What do you do with your data?

Start today
Grow as you need