New blog post:

Risk analysis is a crucial component of resource forecasting for datacenters. By identifying and evaluating potential risks, datacenter managers can ensure that their operations are prepared to handle any potential challenges that may arise, allowing them to deliver reliable, high-quality service to their customers. This blog post explores some types of risks that affect resource demand forecasting in a datacenter.

https://thecloudytales.com/risk-analysis/

#cloudcomputing #capacityplanning #riskanalysis

@urbandata latent demand. Every time they widen the New Jersey Turnpike, traffic increases proportionately.
It works for forecasting IT capacity, too.
#capacityplanning #latentdemand
The only ways to cope with overly-committed resources are
1. Apply architectural constraints (HOV lanes or QoS limitations)
2. Constrain new tasks/users/cars (throttle new users until available capacity returns)

Andrew Newdigate at #PromCon "Tamland: How GitLab.Com Uses Long-Term Monitoring Data For Capacity Forecasting" timestamped at https://www.youtube.com/watch?v=RDdlDRkuVL8&t=6h16m35s

Slides: https://docs.google.com/presentation/d/1jjyIEmVAhU9KN2KqyqEv8N-xHbRPudrWAMMMNmz2PUo/edit?usp=sharing

#monitoring #observability #forecasting #capacityplanning