SCALE API Metrics

This project provides automated tools for collecting performance and health metrics from a SCALE Computing HC3 cluster via its REST API. Metrics are ingested into InfluxDB for real-time analysis, visualization, or alerting. I've also included some other sample scripts such as get_nodeUsage.py, which has been updated to use session-based login instead of Basic Auth.

Overview

• get_stats.py: Collects per-VM metrics (CPU, disk I/O, network usage).
• get_ClusterStats.py: Collects per-node metrics (CPU, memory, drive health and temperature).
• get_stats-basicauth.py / get_ClusterStats-basicauth.py – Variants of the main metric collectors using Basic Auth instead of session tokens.
• start_metric_fetch.sh: Entrypoint script to handle session refresh and sequential execution of both metric collectors.
• gen_sessionID.py – Generates a session token stored at session/sessionLogin.p.
• kill_sessionID.py – Invalidates the current session.
• run_GenDevSession.sh – Shell wrapper to generate a dev session token.
• run_KillDevSession.sh – Shell wrapper to kill a dev session.
• get_nodeUsage.py – Collects node-level CPU/memory metrics, uses session-based login.
• run_getSnapshotReportAll.sh – Optional utility for fetching snapshot-related data or reports (custom functionality).

Requirements

• Python 3.9+
• InfluxDB 2.x
• Virtual environment set up in vfx/
• SCALE HyperCore API access
• Session token stored in session/sessionLogin.p (generated by gen_sessionID.py)

Installation

# Clone this repo
cd /Projects/
git clone https://github.com/Nels2/Scale-Api-Metrics.git
cd Scale-Api-Metrics

# Set up virtual environment
python3 -m venv vfx
source vfx/bin/activate
pip install -r requirements.txt

Usage

To start metric collection manually:

bash start_metric_fetch.sh

This script:

• Checks if your session token is older than 12 hours
• Kills and regenerates session ID if needed
• Runs get_stats.py (VM metrics)
• Runs get_ClusterStats.py (Node metrics)

Output

Metrics are written to InfluxDB:

• VM stats (measurement: vm_metrics)
• VM disk stats (measurement: disk_stats)
• Node CPU (measurement: cpu)
• Node memory (measurement: memory)
• Disk health/temperature (measurement: disk)

Notes

• SSL verification is disabled, you can just remove the flag if you do need it.
• Requires a running InfluxDB instance and pre-created bucket/org, or you can swap out with an alternative.
• Ensure SCALE REST API is accessible via hostname/IP set in host variable in both scripts

Session Management

• Sessions are stored in: session/sessionLogin.p
• If older than 12 hours, a new session is generated automatically

Cron Example

To run every two minutes, add this to your crontab:

*/2 * * * * /bin/bash /Projects/scale_metrics/start_metric_fetch.sh

License: MIT Feel free to use in your own projects.