LDMS Workgroups

The LDMS Users Group runs seasonal sub-workgroups focused on specific technical topics. This page tracks all workgroups and links to their discussion documents.

Requesting document access: click the Drive link for a workgroup and use Drive's Request Access button. If that doesn't resolve, email Nichamon (nichamon@ogc.us).

This page is updated as workgroups start, pause, resume, or wrap up.

---

List of all LDMS Workgroups

Active

Workgroup	Last active
None currently active

Paused

Workgroup	Last active
Multi-Tenant Security	Apr 2026

Completed

Workgroup	Last active
Curated Metrics	2025
YAML Configuration Support	2024
Stream Security	2023
Best Practices	2023
Multi-Tenancy	2023

---

Workgroup summaries

Multi-Tenant Security

This workgroup addresses security and authentication challenges for LDMS in containerized, multi-tenant environments — specifically certificate-based authentication and Unix domain sockets for identity verification and access control. Participants are researching these approaches ahead of the next meeting.

Notes: First round of input gathered. Distinct from Multi-Tenancy — this group is scoped specifically to security/auth, not general resource monitoring.

Materials: https://drive.google.com/drive/folders/12vfVQdaavmI10Bji27EUp6Kla_z7X-Gl?usp=sharing

---

Curated Metrics

This workgroup produced a curated list of LDMS metrics for monitoring system performance — CPU, I/O, network, memory, and file system activity.

Notes: The curated metrics plugin has been merged into the main branch and is slated for the 4.6 release

Materials: https://drive.google.com/drive/folders/1VIDVqRnqCufCw5ZLL6XyWwJIoLzI5vNe?usp=sharing

---

YAML Configuration Support

This workgroup developed YAML-based configuration support for LDMSD within Maestro — consolidating configuration mechanisms, building a dedicated YAML parsing module, and enabling LDMSD to read generated configurations via a new -y command-line option. Work included refining syntax, addressing unsupported commands, and improving sampler error handling, with the parser eventually integrated directly into the LDMS tree.

Notes: See the -y ldmsd command-line option for YAML configuration support, documented on ReadTheDocs and in the ldmsd man page.

Materials: https://drive.google.com/drive/folders/1bzQ-1TEUXBwEKgezWRqBA0qY6CIsJ2S6?usp=sharing

---

Multi-Tenancy

This workgroup looked at strategies for implementing multi-tenancy in LDMS, focused on identifying per-job/user resource utilization. Key themes: tenant-specific plugins, tagging metric data for analysis, and monitoring limitations for resources like GPUs and networks. The group also evaluated scheduling integrations, security, and data partitioning requirements for monitoring concurrent workloads on shared compute nodes.

Notes: Distinct from Multi-Tenant Security — this group covered general per-job/user resource monitoring, not security/auth.

Materials: https://drive.google.com/drive/folders/1h9cWycYCZVLc2TXPfVTEQ2Hz2fgZXgR9?usp=sharing

---

Stream Security

This workgroup covered LDMS Stream v4.5 development — security requirements, API updates, and resource efficiency. Key efforts: implementing ACLs, refining naming conventions, and routing improvements to replace broadcasting. The group met regularly to work through authorization mechanisms and potential security gaps in data-access restrictions within the stream architecture.

Notes: The stream codebase has since diverged from the base assumptions this group worked from — read these notes as historical context, not a current design reference.

Materials: https://drive.google.com/drive/folders/1DMw2LUkkl2zdEryjg7CyN_MPc7ZSIU8W?usp=sharing

---

Best Practices

This workgroup covered LDMS best practices for configuration, deployment, and packaging: normalizing LDMSD configuration formats, transitioning toward declarative YAML-based management, and improving containerized and RPM distribution. It also looked at operational standards for monitoring, configuration tooling (Maestro, Ansible), and sampler plugin integration to improve cluster health diagnostics and management efficiency.

Materials: https://drive.google.com/drive/folders/1ZrRd5XCzgGE4B8rwZS0KqdqGRh2qT4BH?usp=sharing

---