Day 4: Multi-Container Pods
Goal: You build pods with init containers, sidecars, adapters and ambassadors and target individual containers precisely. · Duration: approx. 60 min · Prerequisite: Day 3
Briefing
CKAD: Application Design and Build (20 %) – “Understand
multi-container Pod design patterns (e.g. sidecar, init and others)”.
All containers of a pod share the network (localhost) and
volumes, but run as separate processes with their own logs.
Pod ────────────────────────────────────────────────────────────
| init-1 -> init-2 -> [native sidecar] -> app + classic sidecar |
| one after runs until shared emptyDir, |
| another, each the very end localhost |
| to completion |
─────────────────────────────────────────────────────────────────
Patterns: Init (prepare), Sidecar (extend, e.g. ship logs), Adapter (reshape output), Ambassador (broker connections to the outside).
Drill 1: Init containers
k create ns tag04
k config set-context --current --namespace=tag04
mkdir -p ~/ckad/tag04 && cd ~/ckad/tag04
cat > init.yaml <<'EOF'
apiVersion: v1
kind: Pod
metadata:
name: web
spec:
initContainers:
- name: wait-db
image: busybox:1.37
command:
- sh
- -c
- until nslookup db.tag04.svc.cluster.local; do echo warte auf db; sleep 2; done
- name: content
image: busybox:1.37
command: ["sh", "-c", "echo 'Hallo aus dem Init-Container' > /work/index.html"]
volumeMounts:
- name: html
mountPath: /work
containers:
- name: nginx
image: nginx:1.29
volumeMounts:
- name: html
mountPath: /usr/share/nginx/html
volumes:
- name: html
emptyDir: {}
EOF
k apply -f init.yaml
k get pod web
k logs web -c wait-db --tail=2Expected:
pod/web created
NAME READY STATUS RESTARTS AGE
web 0/1 Init:0/2 0 8s
warte auf db
Init:0/2: neither of the two init containers has
finished. Create the Service that wait-db is waiting
for:
k create service clusterip db --tcp=5432:5432
k get pod web
curl -s $(k get pod web -o jsonpath='{.status.podIP}')Expected:
service/db created
NAME READY STATUS RESTARTS AGE
web 1/1 Running 0 17s
Hallo aus dem Init-Container
Warning:
nslookup dbwith the short name ends with exit code 1 in busybox 1.37, even when the name resolves – the search domains return NXDOMAIN. So use the full name in the init container.
Save your progress? Sign up for free
Drill 2: Inspect init containers
J='{range .status.initContainerStatuses[*]}'
J+='{.name}{" "}{.state.terminated.reason}{"\n"}{end}'
k get pod web -o jsonpath="$J"
k logs web | head -1Expected:
wait-db Completed
content Completed
Defaulted container "nginx" out of: nginx, wait-db (init), content (init)
Without -c, kubectl takes the first regular container
and tells you so. While an init container is running,
k logs web fails with “is waiting to start:
PodInitializing” – then target it with -c wait-db.
Save your progress? Sign up for free
Drill 3: Classic sidecar with emptyDir
The application writes to a file, the sidecar prints it to stdout so
that k logs can see it:
cat > logger.yaml <<'EOF'
apiVersion: v1
kind: Pod
metadata:
name: logger
spec:
containers:
- name: app
image: busybox:1.37
command:
- sh
- -c
- while true; do date '+%T GET /notes 200' >> /log/access.log; sleep 5; done
volumeMounts:
- name: logs
mountPath: /log
- name: log-shipper
image: busybox:1.37
command: ["tail", "-n+1", "-F", "/log/access.log"]
volumeMounts:
- name: logs
mountPath: /log
readOnly: true
volumes:
- name: logs
emptyDir: {}
EOF
k apply -f logger.yaml
k get pod logger
k logs logger -c log-shipper --tail=2Expected:
pod/logger created
NAME READY STATUS RESTARTS AGE
logger 2/2 Running 0 14s
15:48:51 GET /notes 200
15:48:56 GET /notes 200
Save your progress? Sign up for free
Drill 4: Into individual containers
k exec logger -c app -- ls -l /log
k exec logger -c log-shipper -- sh -c 'echo x >> /log/access.log'
k logs logger --all-containers --prefix --tail=1Expected:
total 4
-rw-r--r-- 1 root root 72 Nov 18 15:48 access.log
sh: can't create /log/access.log: Read-only file system
command terminated with exit code 1
[pod/logger/log-shipper] 15:48:56 GET /notes 200
Same volume, different permissions: readOnly: true
applies per volumeMount, i.e. per container.
app only writes to the file, which is why
--all-containers only returns lines from the sidecar.
Save your progress? Sign up for free
Drill 5: Adapter
An adapter reshapes the application’s output into a format another
system expects – here key=value instead of free text. Pods
are immutable, so add a third container under containers:
in logger.yaml:
- name: adapter
image: busybox:1.37
command:
- sh
- -c
- |
tail -n+1 -F /log/access.log \
| awk '{print "time=" $1, "path=" $3, "status=" $4; fflush()}'
volumeMounts:
- name: logs
mountPath: /log
readOnly: truek replace -f logger.yaml $now
k get pod logger
k logs logger -c adapter --tail=2
k get pod logger -o jsonpath='{.spec.containers[*].name}{"\n"}'Expected (shortened):
pod "logger" force deleted from tag04 namespace
pod/logger replaced
NAME READY STATUS RESTARTS AGE
logger 3/3 Running 0 21s
time=15:51:38 path=/notes status=200
time=15:51:43 path=/notes status=200
app log-shipper adapter
Why fflush()? When awk writes to a pipe instead of a
terminal, it buffers – without a flush you see nothing in
k logs for a long time.
Save your progress? Sign up for free
Drill 6: Ambassador
An ambassador takes care of the application’s connection to the
outside. Here the app only talks to localhost:8001;
kubectl proxy in the ambassador handles the API server,
token and TLS:
cat > amb.yaml <<'EOF'
apiVersion: v1
kind: Pod
metadata:
name: client
spec:
containers:
- name: app
image: busybox:1.37
command:
- sh
- -c
- while true; do wget -qO- -T 2 localhost:8001/version | grep gitV; sleep 10; done
- name: api-ambassador
image: registry.k8s.io/kubectl:v1.37.2
args: ["proxy", "--port=8001"]
EOF
k apply -f amb.yaml
k logs client -c app
k logs client -c api-ambassadorExpected:
pod/client created
wget: can't connect to remote host (127.0.0.1): Connection refused
"gitVersion": "v1.37.2",
Starting to serve on 127.0.0.1:8001
The first request goes nowhere: both containers start at the same time, and the proxy was not ready yet. Drill 7 solves that.
Save your progress? Sign up for free
Drill 7: Native sidecar
An init container with restartPolicy: Always is a native
sidecar: it starts before the regular containers (in list order) but
keeps running until the end. Move the ambassador in
amb.yaml:
cat > amb.yaml <<'EOF'
apiVersion: v1
kind: Pod
metadata:
name: client
spec:
initContainers:
- name: api-ambassador
image: registry.k8s.io/kubectl:v1.37.2
args: ["proxy", "--port=8001"]
restartPolicy: Always
containers:
- name: app
image: busybox:1.37
command:
- sh
- -c
- while true; do wget -qO- -T 2 localhost:8001/version | grep gitV; sleep 10; done
EOF
k replace -f amb.yaml $now
k get pod client
k logs client -c appExpected (shortened):
pod/client replaced
NAME READY STATUS RESTARTS AGE
client 2/2 Running 0 15s
"gitVersion": "v1.37.2",
"gitVersion": "v1.37.2",
READY 2/2 counts the native sidecar too. If you need a
real readiness guarantee, give the sidecar a startupProbe –
the regular containers only start once it succeeds.
Warning: A
tcpSocketprobe on port 8001 fails here with “connect: connection refused”: the kubelet checks the pod IP, butkubectl proxyonly listens on127.0.0.1.
Save your progress? Sign up for free
Drill 8: Shutdown order
The real benefit of native sidecars shows with Jobs: once the application is done, the kubelet terminates the sidecars – a classic sidecar would keep the Job running forever.
cat > export.yaml <<'EOF'
apiVersion: batch/v1
kind: Job
metadata:
name: export
spec:
template:
spec:
restartPolicy: Never
initContainers:
- name: prep
image: busybox:1.37
command: ["sh", "-c", "echo prep fertig; sleep 2"]
- name: log-shipper
image: busybox:1.37
restartPolicy: Always
command:
- sh
- -c
- trap 'exit 0' TERM; tail -n+1 -F /log/job.log & wait
volumeMounts:
- name: logs
mountPath: /log
containers:
- name: app
image: busybox:1.37
command:
- sh
- -c
- for i in 1 2 3; do echo Export Teil $i >> /log/job.log; sleep 3; done
volumeMounts:
- name: logs
mountPath: /log
volumes:
- name: logs
emptyDir: {}
EOF
k apply -f export.yaml
k wait --for=condition=complete job/export --timeout=60s
k get job exportExpected:
job.batch/export created
job.batch/export condition met
NAME STATUS COMPLETIONS DURATION AGE
export Complete 1/1 18s 18s
Start and end times of all containers:
S='{.name} {.state.terminated.startedAt} {.state.terminated.finishedAt} '
S="$S"'{.state.terminated.exitCode}{"\n"}'
P=$(k get pods -l job-name=export -o name)
k get $P -o jsonpath="{range .status.initContainerStatuses[*]}$S{end}"
k get $P -o jsonpath="{range .status.containerStatuses[*]}$S{end}"Expected:
prep 2026-11-18T15:50:17Z 2026-11-18T15:50:19Z 0
log-shipper 2026-11-18T15:50:19Z 2026-11-18T15:50:29Z 0
app 2026-11-18T15:50:20Z 2026-11-18T15:50:29Z 0
Startup: prep to completion, then the sidecar, then
app. Shutdown: first app, then the sidecar
(multiple sidecars in reverse order). Why the trap? A shell
as PID 1 ignores SIGTERM. Without it, the kubelet waits out the
30-second grace period and kills the sidecar with exit code 137 – in the
test, the Job took 49 instead of 18 seconds.
Save your progress? Sign up for free
Exercise: Speed Drills
- Pod
twowith containerweb(nginx:1.29) and containersh(busybox:1.37,sleep 3600). Target time: 90 s - Show the last two lines of the container
adapterin podlogger. Target time: 15 s - List the directory
/login the containerlog-shipperoflogger. Target time: 15 s
Solution
k run two --image=nginx:1.29 $do > two.yaml
vim two.yamlChange the container section like this:
containers:
- image: nginx:1.29
name: web
- image: busybox:1.37
name: sh
command: ["sleep", "3600"]k apply -f two.yaml
k logs logger -c adapter --tail=2
k exec logger -c log-shipper -- ls /logExpected (shortened):
pod/two created
time=15:52:28 path=/notes status=200
time=15:52:33 path=/notes status=200
access.log
Save your progress? Sign up for free
Checkpoint
k get pod two -o jsonpath='{.spec.containers[*].name}'Expected: web sh
k get pod client -o jsonpath='{.spec.initContainers[0].restartPolicy}'Expected: Always
k get pod loggerExpected: READY 3/3,
STATUS Running
Challenge
A Deployment notes with notes-api:1.0
should not receive the API key as an environment variable: an init
container writes it to an in-memory emptyDir, and notes-api reads it via
NOTES_API_KEY_FILE.
Hint: emptyDir.medium: Memory creates
the volume as a tmpfs.
Solution
cat > notes.yaml <<'EOF'
apiVersion: apps/v1
kind: Deployment
metadata:
labels:
app: notes
name: notes
spec:
replicas: 1
selector:
matchLabels:
app: notes
template:
metadata:
labels:
app: notes
spec:
initContainers:
- name: write-key
image: busybox:1.37
command: ["sh", "-c", "echo lab-key-123 > /secrets/api-key"]
volumeMounts:
- name: secrets
mountPath: /secrets
containers:
- image: 172.16.0.126:30500/notes-api:1.0
name: notes-api
env:
- name: NOTES_API_KEY_FILE
value: /secrets/api-key
volumeMounts:
- name: secrets
mountPath: /secrets
readOnly: true
volumes:
- name: secrets
emptyDir:
medium: Memory
EOF
k apply -f notes.yaml
k rollout status deployment/notes
k logs deployment/notes | head -1Expected (shortened):
deployment.apps/notes created
deployment "notes" successfully rolled out
Defaulted container "notes-api" out of: notes-api, write-key (init)
time=2026-11-18T15:52:39.722Z level=INFO msg="api key loaded" source=file
source=file confirms the file route. A plaintext key in
the init container is, of course, only a lab value – on Day 6 it comes
from a Secret.
Clean up:
k config set-context --current --namespace=default
k delete ns tag04Save your progress? Sign up for free
Exam Tip
- Native sidecars go under
initContainerswithrestartPolicy: Always– not undercontainers. In the docs under “Sidecar Containers” (Concepts → Workloads → Pods). - With multiple containers, always use
-c <name>withk logsandk exec; container names come fromk get pod <p> -o jsonpath='{.spec.containers[*].name}'. - You cannot add a container to a running pod: change the file and run
k replace -f … $now. For Deployments,k applyis enough.
Enterprise Note
- Service meshes (Istio, Linkerd) inject proxy sidecars automatically via an admission webhook; Istio now uses native sidecars for this.
- Log sidecars are expensive (one container per pod). More common is a DaemonSet agent (Fluent Bit, Vector) that reads the nodes’ container logs.
- Every sidecar needs requests and limits – they count toward the pod and therefore toward quotas and scheduling (Day 7).