This will be a hands-on tutorial where we will have attendees login to containers with LDMS and SOS installed and go through a series of exercises.
exercise 1: Configuration and running samplers
exercise 2: Configure aggregators
exercise 3: Storing Data in CSV Format
Pre-requisites to use LDMS in the containerized environment
Deployment of the containers with pre-built LDMSD, DSOS, SOS, Maestro, and Grafana
Container and network configuration
Accessing data collected by LDMSD in the containers
This tutorial will cover different load balancing options with Maestro as well as an in depth look at cluster configuration. Users will be able to deploy a small scale cluster with Maestro, and work with configuring/deploying different samplers.
Brief overview of commands maestro/maestro_ctrl
Configuration walkthrough
Configuration examples
This tutorial gives a brief overview of an analysis and visualization pipeline, application and system data co-collection, real-world use cases on production HPC systems, and a live demonstration of LDMS analysis examples.
This tutorial will cover how to define a database independent mapping of figures of merit from an LDMS metric set to one or more database row-column definitions.
Practical methods for selecting the number of aggregators, aggregation levels, and storage bandwidth required for a large-scale system given the number of monitored components and metric sets.
This tutorial will explore the efficient distribution of LDMS Metric set data on the Kafka bus.
Demonstrate the client reception of the various available kafka topics
Demonstrate the client query of the available schema/subjects from the Schema Registry