DAOS
daos.io
- Free plan
- Linux
- API
- Self-hosted
What DAOS is
DAOS is a free, open-source, software-defined storage platform for AI and high-performance computing workloads. Its main interface is a scalable key-value-array store exposed through libdaos. POSIX namespaces, HDF5 datasets, Python, MPI-IO, S3 API, and other models can be built over that foundation. Storage pools may use some or all storage engines, while containers hold datasets that can be snapshotted. Fine-grained access controls apply to pools and containers. Objects may use erasure coding, replication, or no protection, and self-healing rebuilds can start when an engine stops responding. DAOS distributes filesystem data and metadata across storage engines and can use RDMA through libfabric or UCX. It supports Linux on x86_64, Kubernetes, hybrid deployment, and S3-compatible access. DAOS 2.8 lists specific SLES, openSUSE Leap, and EL distributions, but not Ubuntu, Fedora, CentOS Stream, or RHEL 8 clones.
Who it suits
DAOS suits teams running AI or HPC workloads that need scalable storage, snapshots, access controls, and configurable data protection. It is intended for Linux-based, self-hosted or hybrid deployments.
Good to know, and what to watch
- Free and open-source storage platform
- Key-value, POSIX, HDF5, MPI-IO, and S3 interfaces
- Erasure coding and replication options
- Fine-grained pool and container ACLs
- Self-healing rebuilds for unresponsive engines
- Linux on x86_64 is required
- DAOS 2.8 excludes Ubuntu and Fedora
- Also excludes CentOS Stream and RHEL 8 clones
- No free trial is listed
The verdict
DAOS combines a scalable storage interface with multiple data models, snapshots, access controls, and protection choices. Check its supported Linux distributions carefully before deployment.
What the maker publishes
- Access control
- Fine-grained ACLs can be set on storage pools and individual containers. source
- Audience
- The site presents DAOS as an open-source storage platform for AI and HPC workloads. source
- Data model
- Its core interface is a scalable key-value-array store, with data models including POSIX namespaces and HDF5 datasets built over it. source
- Data models
- The overview lists POSIX namespaces, HDF5 datasets, and Python among data models built over libdaos. source
- Data protection
- Objects can use erasure coding, replication, or no data protection; DAOS can trigger self-healing rebuilds when an engine becomes unresponsive. source
- Datasets and snapshots
- DAOS containers represent datasets, can be snapshotted, and support fine-grained access control. source
- Deployments
- DAOS can be deployed on premises or in cloud environments. source
- Framework support
- The DAOS v2.8 documentation lists support topics for MPI-IO, HDF5, Python, PyTorch, Spark and Hadoop, and TensorFlow-IO. source
- Integrations
- The DAOS client ecosystem includes libdfs for POSIX namespaces, and its user guide documents a PyTorch interface for loading data from DAOS. source
- License
- The DAOS GitHub repository identifies the software license as BSD-2-Clause Plus Patent License. source
- License
- The project repository displays the BSD-2-Clause Plus Patent License. source
- Network support
- DAOS can use libfabric or UCX and lists Ethernet, RoCE, InfiniBand, Omni-Path, and Slingshot fabrics. source
- Notable limits
- DAOS 2.8 does not support Ubuntu, Fedora, CentOS Stream, or RHEL 8 and its clones. source
- Performance
- DAOS distributes filesystem data and metadata across storage engines and can use RDMA through libfabric or UCX when available. source
- Performance
- The site describes distributed metadata, ultra-high bandwidth and IOPS, and time-to-first-byte below 20μs when using RDMA where available. source
- Protection and recovery
- Objects can use erasure coding or replication, and DAOS can trigger self-healing rebuilds when an engine becomes unresponsive. source
- Purpose
- DAOS is an open-source, software-defined storage system for AI and HPC workloads. source
- Purpose
- DAOS is an open-source, software-defined storage system for AI and HPC workloads. source
- Release
- The DAOS documentation identifies its current documentation set as version 2.8. source
- Security
- DAOS supports access control lists on pools and individual containers. source
- Snapshots and scale
- DAOS containers are user-managed datasets that can be snapshotted; the site says a POSIX container can hold trillions of files or directories and petabytes of data. source
- Storage interface
- Its baseline interface is a high-performance, scalable key-value-array store exposed through libdaos. source
- Storage pools
- Pools can span all or some storage engines, and pool access can be controlled with fine-grained ACLs. source
- Support
- The DAOS site lists member support by email and links to a support center and Jira bug tracker. source
Checked against the maker's pages on 3 Oct 2026.