Category: Tech Notes
-
Capacity Forecasting From Infrastructure Telemetry (Infra to Data Science Series, Part 22)
A straight line on a seasonal metric tells you when the average crosses your limit, not when the weekly peak does. Forecast with Holt-Winters, read the date off the upper prediction band, and plan against the risk you can afford.
-
Anomaly Detection on Metrics and Time Series for Infra Telemetry (Infra to Data Science Series, Part 21)
Static thresholds on seasonal infrastructure metrics measure the time of day, not trouble. Deseasonalise with STL, score the residual with a median and MAD based z, and reach for an isolation forest when several metrics move together.
-
Experiment Tracking and Model Registry for Infra Data (Infra to Data Science Series, Part 20)
Log every training run, register the good ones as immutable versions, and let a single champion alias decide what serves, so promotion and rollback each become one line. Built on the incident classifier, with the real failures that bite.
-
Pipelines and CI/CD for Machine Learning on Infra Data (Infra to Data Science Series, Part 19)
A pipeline and a CI gate that refuse to promote a model unless it clears a metric on a time aware split, built on the incident classifier from earlier parts, with the real failures that break each stage.
-
Monitoring Models in Production for Drift and Decay (Infra to Data Science Series, Part 18)
A served model decays quietly. Here is how to catch it with input and prediction drift checks, a KS test, and a PSI threshold you can page on, all on your own telemetry.
-
Serving a Model Batch and Real Time for Infrastructure Engineers (Infra to Data Science Series, Part 17)
Your trained model becomes two deployables, a batch scoring job on a schedule and a real time endpoint. Here is how an infrastructure engineer serves both, and which one most infra work actually needs.
-
MLOps for Infrastructure Engineers, the Operations You Already Run (Infra to Data Science Series, Part 16)
Most of MLOps is the operational discipline infrastructure engineers already have. This part maps your runbooks, CI and monitoring onto a model, and names the two or three things that are genuinely new.
-
Feature Engineering on Operational Data for Infra Telemetry (Infra to Data Science Series, Part 15)
One lag feature took a fair 0.879 to 0.934 on a month of infra telemetry, while a rate of change added nothing. How to build past only lag, rolling, standard deviation and EWMA features that lift a score without leaking.
-
Model Evaluation Without Fooling Yourself (Infra to Data Science Series, Part 14)
A single new feature lifted this model AUC from 0.889 to 0.975 on infra telemetry, and none of it was real. How to catch leakage in features, preprocessing and folds, and read a score you can defend.
-
Your First Model, From Metric Threshold to Trained Classifier (Infra to Data Science Series, Part 13)
Your production alert rule is a one feature classifier, and it probably catches almost nothing. Here is how to measure it honestly and beat it with a trained model whose operating point you actually choose.
-
Machine Learning Fundamentals for Infra Engineers (Infra to Data Science Series, Part 12)
Machine learning fundamentals for infra engineers, using your own incident labels: why accuracy lies on rare events, and how a baseline, recall and precision decide a real model.
-
Probability and Distributions for Infra Telemetry (Infra to Data Science Series, Part 11)
Fit named distributions to your own latency and arrival data in Python, and see exactly where a clean fitted curve understates the tail that breaches your SLO.
Architect’s Toolkit
PJ’s Tools
- Infra 360 Hub – All Series
- VCF 9 Interactive Walkthroughs
- VCF Design Cheatsheet
- VCF Upgrade Planner
- VCF 9 Series Hub
- VCF Deployment Hub
- AI Stack Hub
- AI Infra Sizing & Cost Calculator
- LLM & RAG Cost Calculator
- DrJhaGPT – Ask Pranay
VMware Cloud Foundation
- VCF Documentation
- VCF 9 Planning & Preparation Workbook
- VCF Bill of Materials (BoM)
- VMware Compatibility Guide
- VMware Interoperability Matrix
- VMware Configuration Maximums
- VMware Ports & Protocols
- VMware Hands-on Labs
- RVTools Download
Nutanix
AI & Cloud-Native Platform
- NVIDIA Build (Model Catalog)
- NVIDIA AI Enterprise Reference Architecture
- NVIDIA NIM Performance Benchmarking
- NVIDIA NGC Catalog
- NeMo Microservices Helm Chart
- Helm Charts Repository
- Hugging Face Models
Architecture & Design
About the Author
Dr Pranay Jha
Dr. Pranay Jha is a Cloud and AI Consultant with 18+ years of experience in hybrid cloud, virtualization, and enterprise infrastructure transformation. He specializes in VMware technologies, multi-cloud strategy, and Generative AI solutions. He holds a PhD in Computer Applications with research focused on Cloud and AI, has published multiple research papers, and has been a VMware vExpert since 2016 and a VMUG Community Leader.
You May Have Missed
DrJha