Hi there,

Please find below the DAOS community newsletter for March 2022.

Past Events (February)

• Dell Technologies HPC Community Event (February 9^th)
  DAOS: High Performance Storage for All HPC Workloads and Scales
  Kelsey Prantis (Intel)
  https://www.dellhpc.org/event_2-9-2022.html

Upcoming Events

• 2022 Energy HPC Conference (March 3^rd)
  DAOS Use at Argonne
  Kevin Harms (Argonne)
  Johann Lombardi (Intel)
  Mohamad Chaarawi (Intel)
  https://energyhpc.rice.edu/program/

Release

• Current stable release is 2.0.1 and 2.0.2 is under development.
• Master is the development branch for the future 2.2 release.
• Major recent 2.0 changes to be available in 2.0.2:
• Build go code shipped in RPM with go >= 1.17
• Fix a bug with key enumeration and multiple shards
• Many documentation fixes (e.g. fix broken links)
• Use larger ABT stack for NVMe poll ULT
• Reject integer keys that are smaller than 8 bytes
• Update the C-based raft implementation to pick a membership fix
• Disable dfuse-level readahead when FUSE is configured in writeback mode and already drive readahead from the kernel.
• Major recent master changes:
• Fix a problem between mercury and libfabric causing some bulk transfers to be stuck indefinitely
• Reduce CPU load caused by the engine when DAOS system is idle
• Add support for owner/group to libdfs
• Automatically calculate the required number of huge pages
• Several VMD and SSD hotplug fixes
• Document the need to disable shared received context for Intel E810-C cards
• Rewrite of the netdetect module
• Add UCX support to the manual (non RPM) build infrastructure
• What is coming:
• 2.0.2 validation
• 2.2.0 feature freeze

R&D

• Major features under development:
• Checksum scrubbing
• UCX support
• LDMS plugin to export DAOS metrics (targeted for 2.2)
• API to collect libdaos metrics to be integrated with Darshan (targeted for 2.4)
• Multi-user dfuse (targeted for 2.4)
• More aggressive caching in dfuse for AI APPs (targeted for 2.4)
• TensorFlow-IO plugin for DFS
• https://github.com/daos-stack/tensorflow-io-daos/blob/devel/docs/daos_tf_docs.md
• Pathfinding:
• MariaDB DAOS engine with predicate pushdown to the DAOS storage nodes
• Prototyped DAOS MariaDB engine available here: https://github.com/daos-stack/mariadb
• PR for pipeline API (#6238)
• Work in progress to support pipeline API in the engine
• Leveraging the Intel Data Streaming Accelerator (DSA) to accelerate DAOS
• Prototype leveraging DSA for VOS aggregation delivered
• S3 support via a DAOS backend to Rados Gateway (RGW)
• GPU data path optimizations

News

• Sodacode 2022 hackathon started today with participation of the DAOS team.
  The list of DAOS activities offered in this hackathon has been created in jira: https://daosio.atlassian.net/browse/DAOS-10010?jql=labels%20%3D%20%22SODACODE2022%22%20and%20statusCategory%20!%3D%20Done

See https://events.linuxfoundation.org/sodacode/ for more information.