K8S Notes
Bookmark this to keep an eye on my K8s notes updates!
Project maintained by kevinsulatra Hosted on GitHub Pages — Theme by mattgraham
Autoscaling
The autoscaling pattern allows for the dynamic adjustment of resources based on the current demand. Typically, this involves horizontal scaling, which means spawning new compute resources like copies of application processes, virtual machines, or even new racks of servers. Another form of scaling is vertical scaling, which refers to changing the size of the underlying hardware.
Autoscaling is often based on metrics like CPU and memory, but other methods such as time or business metrics can also be considered. Configuring autoscaling involves setting minimum and maximum instance limits and defining a triggering metric. Cloud environments with on-demand pricing models are particularly effective for autoscaling, as they can provision resources quickly and even scale to zero when not needed.
Scaling your applications and infrastructure can significantly increase availability and resilience, even in more traditional environments.
