OOMKilled containers in Kubernetes - Step-by-Step Execution

Process Flow - OOMKilled containers

Container starts running

↓

Container uses memory

↓

Memory usage > limit?

No→Continue running

Yes↓

Kernel kills container (OOMKilled)

↓

Container status set to OOMKilled

↓

Pod restarts or fails

This flow shows how a container runs, uses memory, and if it exceeds its memory limit, the system kills it with OOMKilled status.

Execution Sample

Kubernetes

kubectl get pods
kubectl describe pod mypod
kubectl logs mypod
kubectl get events

Commands to check pod status, details, logs, and events to diagnose OOMKilled containers.

Process Table

Step	Action	Memory Usage (MiB)	Memory Limit (MiB)	Result	Container Status
1	Container starts	0	100	Running	Running
2	Memory usage rises	50	100	Running	Running
3	Memory usage rises	90	100	Running	Running
4	Memory usage rises	110	100	Exceeded limit	OOMKilled
5	Kernel kills container	N/A	100	Container killed	OOMKilled
6	Pod restarts or fails	N/A	100	Restarting or Failed	Running

💡 Container memory usage exceeded limit, kernel killed container with OOMKilled status.

Status Tracker

Variable	Start	After Step 2	After Step 3	After Step 4	Final
Memory Usage (MiB)	0	50	90	110	N/A
Container Status	Running	Running	Running	OOMKilled	Running

Key Moments - 3 Insights

Why does the container status change to OOMKilled even though the pod is still listed?

The container is killed by the system because it used more memory than allowed. The pod still exists but the container inside it has stopped with OOMKilled status, as shown in step 4 and 5 of the execution table.

What does 'Memory Limit' mean and why is it important?

Memory Limit is the maximum memory the container can use. If the container uses more than this, the system kills it to protect other containers. This is why in step 4, exceeding 100 MiB causes OOMKilled.

Can the pod restart automatically after OOMKilled?

Yes, depending on the pod's restart policy, it may restart the container after OOMKilled, as shown in step 6. But if it keeps exceeding memory, it may keep failing.

Visual Quiz - 3 Questions

Look at the execution table, what is the container status at step 3?

A Pending

B Running

C OOMKilled

D Failed

💡 Hint

Check the 'Container Status' column at step 3 in the execution table.

At which step does the container exceed its memory limit?

A Step 4

B Step 2

C Step 3

D Step 5

💡 Hint

Look at the 'Memory Usage' and 'Memory Limit' columns to find when usage surpasses limit.

If the memory limit was increased to 120 MiB, what would change in the execution table?

A Container would be killed at step 3

B Container would be OOMKilled at step 4

C Container would continue running past step 4

D No change in container status

💡 Hint

Compare memory usage at step 4 (110 MiB) with new limit 120 MiB.

Concept Snapshot

OOMKilled containers happen when a container uses more memory than its limit.
The system kills the container to protect others.
Check pod status with 'kubectl describe pod' and events.
Adjust memory limits to prevent OOMKilled.
Pod may restart depending on restart policy.

Full Transcript

When a container runs in Kubernetes, it has a memory limit set. As it uses memory, if it stays below the limit, it runs fine. But if it uses more memory than allowed, the system kills it to protect other containers. This kill is called OOMKilled. You can see this status by describing the pod or checking events. The pod may restart the container depending on its restart policy. To avoid OOMKilled, increase memory limits or optimize container memory use.