Grokking DevOps and SRE Interviews
Vote

0% completed

​

How Kubernetes Schedules Pods

A node has 4 CPUs and 16 GiB of memory. After the kubelet reserves room for the operating system, its own processes, and a safety margin, this node offers 3,800 millicores of CPU and 14 GiB of memory to pods. One millicore is a thousandth of a CPU, so 500 millicores, written 500m, is half a CPU.

A team's pods each request 500m of CPU and 3 GiB of memory. By CPU, the node fits 3,800 / 500 = 7.6, so 7 pods. By memory, it fits 14 / 3 = 4.67, so 4 pods. The smaller number wins, and the fifth pod waits even though 1,800 millicores of CPU remain unreserved.

.....

.....

.....

Like the course? Get enrolled and start learning!

Reading Progress

0%


Vote for new content