Probe failure and container restart behavior in Kubernetes - Step-by-Step Execution

Process Flow - Probe failure and container restart behavior

Pod starts container

↓

Kubelet runs probes

↓

Probe success?

No→Failure count increments

↓

Failure count >= threshold?

↓

Container runs normally

↓

Failure count >= threshold?

Yes→Container restart triggered

No↓

Continue monitoring probes

This flow shows how Kubernetes runs health probes on containers, counts failures, and restarts containers when failures reach a limit.

Execution Sample

Kubernetes

livenessProbe:
  httpGet:
    path: /healthz
    port: 8080
  failureThreshold: 3
  periodSeconds: 5

Defines a liveness probe that checks /healthz every 5 seconds and restarts container after 3 failures.

Process Table

Check Number	Probe Result	Failure Count	Restart Triggered	Action
1	Success	0	No	Container continues running
2	Failure	1	No	Monitor continues
3	Failure	2	No	Monitor continues
4	Failure	3	Yes	Container restart triggered
5	Success	0	No	Container runs after restart

💡 After 3 consecutive failures, container restarts and failure count resets

Status Tracker

Variable	Start	After Check 1	After Check 2	After Check 3	After Check 4	After Check 5
failureCount	0	0	1	2	3	0
containerState	Running	Running	Running	Running	Restarting	Running

Key Moments - 3 Insights

Why does the failure count reset after the container restarts?

Because when the container restarts (see row 4 in execution_table), Kubernetes resets the failure count to zero to start fresh monitoring.

Does a single probe failure immediately restart the container?

No, the container restarts only after failureCount reaches the failureThreshold (3 in this example), as shown in rows 2-4.

What happens if a probe succeeds after some failures?

The failure count resets to zero (see row 5), so the container is considered healthy again.

Visual Quiz - 3 Questions

Look at the execution table, what is the failureCount after the 3rd probe check?

A 1

B 2

C 3

D 0

💡 Hint

Check the 'Failure Count' column at row 3 in the execution_table.

At which check number does the container restart get triggered?

A 2

B 3

C 4

D 5

💡 Hint

Look at the 'Restart Triggered' column in the execution_table.

If the failureThreshold was set to 2, when would the container restart occur?

A After check 2

B After check 3

C After check 4

D After check 5

💡 Hint

Compare failureCount with failureThreshold in the execution_table rows.

Concept Snapshot

Kubernetes probes check container health periodically.
Failures increment a failureCount.
When failureCount reaches failureThreshold, container restarts.
Success resets failureCount to zero.
This ensures unhealthy containers get restarted automatically.

Full Transcript

Kubernetes uses probes like livenessProbe to check if a container is healthy. Each probe runs periodically. If a probe fails, Kubernetes increases a failure count. When this count reaches a set threshold, Kubernetes restarts the container to fix issues. If a probe succeeds, the failure count resets to zero. This cycle helps keep containers running smoothly by restarting them only after repeated failures, not just one failure.