Startup probe concept in Kubernetes - Step-by-Step Execution

Process Flow - Startup probe concept

Pod starts

↓

Startup probe runs

↓

Probe success?

No→Retry after delay

↓

Failure threshold reached?

No→Retry

↓

Container marked started

↓

Readiness & Liveness probes start

The startup probe runs repeatedly after pod start until it succeeds or fails too many times. Success means container is ready to run readiness and liveness probes. Failure triggers pod restart.

Execution Sample

Kubernetes

startupProbe:
  exec:
    command: ["cat", "/tmp/healthy"]
  failureThreshold: 3
  periodSeconds: 5

This startup probe runs a command every 5 seconds, retries up to 3 times before failing.

Process Table

Attempt	Time (s)	Command Result	Probe Status	Action
1	0	File not found	Failure	Retry after 5s
2	5	File not found	Failure	Retry after 5s
3	10	File not found	Failure	Failure threshold reached, pod restart
4	15	Pod restarts	N/A	Startup probe restarts

💡 Failure threshold 3 reached at 10s, pod restarts and startup probe restarts

Status Tracker

Variable	Start	After Attempt 1	After Attempt 2	After Attempt 3	After Restart
failureCount	0	1	2	3	0
probeStatus	N/A	Failure	Failure	Failure	N/A
podState	Starting	Starting	Starting	Restarting	Starting

Key Moments - 3 Insights

Why does the pod restart after 3 failed attempts?

Because failureThreshold is 3, after 3 consecutive failures (see execution_table rows 1-3), Kubernetes restarts the pod to try again.

When do readiness and liveness probes start running?

They start only after the startup probe succeeds, which means the container is fully started (not shown in this failure example).

What happens to failureCount after pod restart?

It resets to 0 because the pod and its probes start fresh (see variable_tracker After Restart).

Visual Quiz - 3 Questions

Look at the execution table, what is the failureCount after the second attempt?

A 3

B 1

C 2

D 0

💡 Hint

Check variable_tracker row for failureCount after Attempt 2

At which attempt does the pod restart happen?

A After Attempt 3

B After Attempt 1

C After Attempt 2

D After Attempt 4

💡 Hint

See execution_table Action column for when failure threshold triggers restart

If the failureThreshold was set to 5, what would change in the execution table?

A Pod would restart after 3 failures

B Pod would restart after 5 failures

C Pod would never restart

D Pod would restart immediately

💡 Hint

failureThreshold controls max failures before restart, see execution_table rows and exit_note

Concept Snapshot

Startup probe runs after pod starts to check if container is ready.
It retries command or check every periodSeconds.
If failures reach failureThreshold, pod restarts.
Success means readiness and liveness probes start.
Useful for slow-starting containers.

Full Transcript

The startup probe in Kubernetes runs repeatedly after a pod starts to check if the container is ready. It executes a command or HTTP check every periodSeconds. If the probe fails, it retries until failureThreshold is reached. If failures reach this threshold, Kubernetes restarts the pod to try again. Once the startup probe succeeds, readiness and liveness probes begin to monitor the container's health. This mechanism helps with containers that take longer to start, preventing premature restarts.