TCP probe configuration in Kubernetes - Step-by-Step Execution

Process Flow - TCP probe configuration

Pod starts

↓

Kubelet sets TCP probe

↓

Kubelet tries TCP connection to pod's port

↓

Pod Ready

↓

Repeat probe at intervals

The kubelet repeatedly tries to open a TCP connection to the pod's port. If successful, the pod is marked ready; if not, it is marked not ready.

Execution Sample

Kubernetes

readinessProbe:
  tcpSocket:
    port: 8080
  initialDelaySeconds: 5
  periodSeconds: 10

This config sets a TCP readiness probe on port 8080, starting 5 seconds after pod start, checking every 10 seconds.

Process Table

Step	Action	TCP Connection Attempt	Result	Pod Status
1	Wait initialDelaySeconds (5s)	No attempt	N/A	Unknown
2	Attempt TCP connection to port 8080	Connect to 8080	Success	Ready
3	Wait periodSeconds (10s)	No attempt	N/A	Ready
4	Attempt TCP connection to port 8080	Connect to 8080	Success	Ready
5	Wait periodSeconds (10s)	No attempt	N/A	Ready
6	Attempt TCP connection to port 8080	Connect to 8080	Failure	Not Ready
7	Wait periodSeconds (10s)	No attempt	N/A	Not Ready
8	Attempt TCP connection to port 8080	Connect to 8080	Success	Ready

💡 Probe continues indefinitely; pod status updates based on TCP connection success or failure.

Status Tracker

Variable	Start	After Step 2	After Step 4	After Step 6	After Step 8
Pod Status	Unknown	Ready	Ready	Not Ready	Ready
TCP Connection Result	N/A	Success	Success	Failure	Success

Key Moments - 2 Insights

Why does the pod status change to Not Ready at step 6 even though the pod is running?

Because the TCP probe at step 6 failed to connect to port 8080, indicating the pod is not responding on that port, so kubelet marks it Not Ready (see execution_table row 6).

What happens during initialDelaySeconds before the first probe?

No TCP connection attempts are made during initialDelaySeconds (step 1), so pod status remains Unknown until the first probe (execution_table row 1).

Visual Quiz - 3 Questions

Look at the execution table, what is the Pod Status after the first successful TCP connection?

A Unknown

B Not Ready

C Ready

D Terminated

💡 Hint

Check execution_table row 2 under Pod Status.

At which step does the TCP connection fail?

A Step 4

B Step 6

C Step 8

D Step 2

💡 Hint

Look at execution_table row 6 under TCP Connection Attempt and Result.

If initialDelaySeconds was set to 0, what would change in the execution table?

A Step 1 would attempt TCP connection immediately

B Pod Status would start as Ready

C TCP connection attempts would stop

D Pod would never become Not Ready

💡 Hint

Refer to execution_table row 1 where no attempt is made due to initialDelaySeconds.

Concept Snapshot

TCP Probe Configuration in Kubernetes:
- Uses tcpSocket to check pod port availability
- initialDelaySeconds delays first check
- periodSeconds sets interval between checks
- Success means pod is Ready; failure means Not Ready
- Probe runs repeatedly to monitor pod health

Full Transcript

In Kubernetes, a TCP probe checks if a pod is ready by trying to open a TCP connection to a specified port. The kubelet waits for an initial delay, then attempts the connection repeatedly at set intervals. If the connection succeeds, the pod is marked Ready; if it fails, the pod is marked Not Ready. This helps Kubernetes know if the pod is healthy and can receive traffic. The example shows a probe on port 8080 with a 5-second initial delay and 10-second intervals. The execution table traces connection attempts and pod status changes over time.