Files
podman/test/system/700-play.bats
Jan Kaluza 224e791161 Replace podman pause image with rootfs.
This commit removes the code to build a local pause
image from the Containerfile. It is replaced with
code to find the catatonit binary and include it in
the Rootfs.

This removes the need to build a local pause container
image.

The same logic is also applied to createServiceContainer
which is originally also based on the pause image.

Fixes: #23292

Signed-off-by: Jan Kaluza <jkaluza@redhat.com>
2025-04-17 08:36:27 +02:00

1287 lines
43 KiB
Bash

#!/usr/bin/env bats -*- bats -*-
#
# Test podman play
#
load helpers
load helpers.network
load helpers.registry
# helper function: writes a yaml file with customizable values
function _write_test_yaml() {
# All of these are available to our caller
TESTYAML=$PODMAN_TMPDIR/test.yaml
PODNAME="p-$(safename)"
CTRNAME="c-$(safename)"
PODCTRNAME="$PODNAME-$CTRNAME"
# Function args must all be of the form 'keyword=value' (value may be null)
local annotations=
local labels="app: test"
local command=""
local image="$IMAGE"
local volume=
for i;do
# This will error on 'foo=' (no value). That's totally OK.
local value=$(expr "$i" : '[^=]*=\(.*\)')
case "$i" in
annotations=*) annotations="$value" ;;
labels=*) labels="$value" ;;
name=*) PODNAME="$value" ;;
command=*) command="$value" ;;
image=*) image="$value" ;;
ctrname=*) CTRNAME="$value" ;;
volume=*) volume="$value" ;;
*) die "_write_test_yaml: cannot grok '$i'" ;;
esac
done
# These three header lines are common to all yamls.
# Note: use >> (append), not > (overwrite), for multi-pod test
cat >>$TESTYAML <<EOF
apiVersion: v1
kind: Pod
metadata:
EOF
if [[ -n "$annotations" ]]; then
echo " annotations:" >>$TESTYAML
echo " $annotations" >>$TESTYAML
fi
if [[ -n "$labels" ]]; then
echo " labels:" >>$TESTYAML
echo " $labels" >>$TESTYAML
fi
if [[ -n "$PODNAME" ]]; then
echo " name: $PODNAME" >>$TESTYAML
fi
# We always have spec and container lines...
echo "spec:" >>$TESTYAML
echo " containers:" >>$TESTYAML
# ...but command is optional. If absent, assume our caller will fill it in.
if [[ -n "$command" ]]; then
cat <<EOF >>$TESTYAML
- command:
- $command
image: $image
name: $CTRNAME
resources: {}
EOF
# only makes sense when command is given
if [[ -n "$volume" ]]; then
cat <<EOF >>$TESTYAML
securityContext:
runAsUser: 1000
runAsGroup: 3000
fsGroup: 2000
allowPrivilegeEscalation: true
capabilities: {}
privileged: false
seLinuxOptions:
level: "s0:c1,c2"
readOnlyRootFilesystem: false
volumeMounts:
- mountPath: /testdir:z
name: home-podman-testdir
workingDir: /
volumes:
- hostPath:
path: $volume
type: Directory
name: home-podman-testdir
EOF
fi
# Done.
echo "status: {}" >>$TESTYAML
fi
# For debugging
echo "# test yaml:"
sed -e "s/^/ /g" <$TESTYAML
}
RELABEL="system_u:object_r:container_file_t:s0"
# bats test_tags=ci:parallel
@test "podman kube with stdin" {
TESTDIR=$PODMAN_TMPDIR/testdir
mkdir -p $TESTDIR
_write_test_yaml command=top volume=$TESTDIR
run_podman kube play - < $TESTYAML
if selinux_enabled; then
run ls -Zd $TESTDIR
is "$output" "${RELABEL} $TESTDIR" "selinux relabel should have happened"
fi
run_podman pod rm -t 0 -f $PODNAME
}
# bats test_tags=ci:parallel
@test "podman play" {
# Testing that the "podman play" cmd still works now that
# "podman kube" is an option.
TESTDIR=$PODMAN_TMPDIR/testdir
mkdir -p $TESTDIR
_write_test_yaml command=top volume=$TESTDIR
run_podman play kube $TESTYAML
if selinux_enabled; then
run ls -Zd $TESTDIR
is "$output" "${RELABEL} $TESTDIR" "selinux relabel should have happened"
fi
# Now rerun twice to make sure nothing gets removed
run_podman 125 play kube $TESTYAML
is "$output" ".* is in use: pod already exists"
run_podman 125 play kube $TESTYAML
is "$output" ".* is in use: pod already exists"
run_podman pod rm -t 0 -f $PODNAME
}
# bats test_tags=ci:parallel
@test "podman play --service-container" {
skip_if_remote "service containers only work locally"
# Create the YAMl file
_write_test_yaml command=top
# Run `play kube` in the background as it will wait for the service
# container to exit.
timeout --foreground -v --kill=10 60 \
$PODMAN play kube --service-container=true --log-driver journald $TESTYAML &>/dev/null &
# Wait for the container to be running
container_a=$PODCTRNAME
container_running=
for i in $(seq 1 20); do
run_podman "?" container wait $container_a --condition="running"
if [[ $status == 0 ]]; then
container_running=1
break
fi
sleep 0.5
# Just for debugging
run_podman ps -a
done
if [[ -z "$container_running" ]]; then
die "container $container_a did not start"
fi
# The name of the service container is predictable: the first 12 characters
# of the hash of the YAML file followed by the "-service" suffix
yaml_sha=$(sha256sum $TESTYAML)
service_container="${yaml_sha:0:12}-service"
# Make sure that the service container exists and runs.
run_podman container inspect $service_container --format "{{.State.Running}}"
is "$output" "true"
run_podman container inspect $service_container --format '{{.Config.StopTimeout}}'
is "$output" "10" "StopTimeout should be initialized to 10"
# Stop the *main* container and make sure that
# 1) The pod transitions to Exited
# 2) The service container is stopped
# #) The service container is marked as a service container
run_podman stop $PODCTRNAME
_ensure_pod_state $PODNAME Exited
_ensure_container_running $service_container false
run_podman container inspect $service_container --format "{{.IsService}}"
is "$output" "true"
# Restart the pod, make sure the service is running again
run_podman pod restart $PODNAME
run_podman container inspect $service_container --format "{{.State.Running}}"
is "$output" "true"
# Check for an error when trying to remove the service container
run_podman 125 container rm $service_container
is "$output" "Error: container .* is the service container of pod(s) .* and cannot be removed without removing the pod(s)"
run_podman 125 container rm --force $service_container
is "$output" "Error: container .* is the service container of pod(s) .* and cannot be removed without removing the pod(s)"
# Kill the pod and make sure the service is not running
run_podman pod kill $PODNAME
_ensure_container_running $service_container false
run_podman network ls
# Remove the pod and make sure the service is removed along with it
run_podman pod rm $PODNAME
run_podman 1 container exists $service_container
}
# bats test_tags=ci:parallel
@test "podman kube --network" {
_write_test_yaml command=top
run_podman kube play --network host $TESTYAML
is "$output" "Pod:.*" "podman kube play should work with --network host"
run_podman pod inspect --format "{{.InfraConfig.HostNetwork}}" $PODNAME
is "$output" "true" ".InfraConfig.HostNetwork"
run_podman pod rm -t 0 -f $PODNAME
if has_slirp4netns; then
run_podman kube play --network slirp4netns:port_handler=slirp4netns $TESTYAML
run_podman pod inspect --format {{.InfraContainerID}} "${lines[1]}"
infraID="$output"
run_podman container inspect --format "{{.HostConfig.NetworkMode}}" $infraID
is "$output" "slirp4netns" "network mode slirp4netns is set for the container"
fi
run_podman pod rm -t 0 -f $PODNAME
run_podman kube play --network none $TESTYAML
run_podman pod inspect --format {{.InfraContainerID}} "${lines[1]}"
infraID="$output"
run_podman container inspect --format "{{.HostConfig.NetworkMode}}" $infraID
is "$output" "none" "network mode none is set for the container"
run_podman container exists $PODCTRNAME
run_podman kube down $TESTYAML
run_podman 1 container exists $PODCTRNAME
}
# bats test_tags=ci:parallel
@test "podman kube play read-only" {
YAML=$PODMAN_TMPDIR/test.yml
podname="p-$(safename)"
c1name="c1-$(safename)"
c2name="c2-$(safename)"
c3name="c3-$(safename)"
# --restart=no is crucial: without that, the "podman wait" below
# will spin for indeterminate time.
run_podman create --pod new:$podname --restart=no --name $c1name $IMAGE touch /testrw
run_podman create --pod $podname --read-only --restart=no --name $c2name $IMAGE touch /testro
run_podman create --pod $podname --read-only --restart=no --name $c3name $IMAGE sh -c "echo "#!echo hi" > /tmp/testtmp; chmod +x /tmp/test/tmp; /tmp/testtmp"
# Generate and run from yaml. (The "cat" is for debugging failures)
run_podman kube generate $podname -f $YAML
cat $YAML
run_podman kube play --replace $YAML
# Wait for all containers and check their exit statuses
run_podman wait ${podname}-${c1name} ${podname}-${c2name} ${podname}-${c3name}
is "${lines[0]}" 0 "exit status: touch /file on read/write container"
is "${lines[1]}" 1 "exit status: touch /file on read-only container"
is "${lines[2]}" 0 "exit status: touch on /tmp is always ok, even on read-only container"
# Confirm config settings
run_podman container inspect --format '{{.HostConfig.ReadonlyRootfs}}' ${podname}-${c1name} ${podname}-${c2name} ${podname}-${c3name}
is "${lines[0]}" "false" "ReadonlyRootfs - container 1"
is "${lines[1]}" "true" "ReadonlyRootfs - container 2"
is "${lines[2]}" "true" "ReadonlyRootfs - container 3"
# Clean up
run_podman kube down - < $YAML
run_podman 1 container exists ${podname}-${c1name}
run_podman 1 container exists ${podname}-${c2name}
run_podman 1 container exists ${podname}-${c3name}
}
# bats test_tags=ci:parallel
@test "podman kube play read-only from containers.conf" {
containersconf=$PODMAN_TMPDIR/containers.conf
cat >$containersconf <<EOF
[containers]
read_only=true
EOF
YAML=$PODMAN_TMPDIR/test.yml
podname="p-$(safename)"
c1name="c1-$(safename)"
c2name="c2-$(safename)"
c3name="c3-$(safename)"
# --restart=no is crucial: without that, the "podman wait" below
# will spin for indeterminate time.
CONTAINERS_CONF_OVERRIDE="$containersconf" run_podman create --pod new:$podname --read-only=false --restart=no --name $c1name $IMAGE touch /testrw
CONTAINERS_CONF_OVERRIDE="$containersconf" run_podman create --pod $podname --restart=no --name $c2name $IMAGE touch /testro
CONTAINERS_CONF_OVERRIDE="$containersconf" run_podman create --pod $podname --restart=no --name $c3name $IMAGE touch /tmp/testtmp
# Inspect settings in created containers
CONTAINERS_CONF_OVERRIDE="$containersconf" run_podman container inspect --format '{{.HostConfig.ReadonlyRootfs}}' $c1name $c2name $c3name
is "${lines[0]}" "false" "ReadonlyRootfs - container 1, created"
is "${lines[1]}" "true" "ReadonlyRootfs - container 2, created"
is "${lines[2]}" "true" "ReadonlyRootfs - container 3, created"
# Now generate and run kube.yaml on a machine without the defaults set
CONTAINERS_CONF_OVERRIDE="$containersconf" run_podman kube generate $podname -f $YAML
cat $YAML
run_podman kube play --replace $YAML
# Wait for all containers and check their exit statuses
run_podman wait ${podname}-${c1name} ${podname}-${c2name} ${podname}-${c3name}
is "${lines[0]}" 0 "exit status: touch /file on read/write container"
is "${lines[1]}" 1 "exit status: touch /file on read-only container"
is "${lines[2]}" 0 "exit status: touch on /tmp is always ok, even on read-only container"
# Confirm settings again
run_podman container inspect --format '{{.HostConfig.ReadonlyRootfs}}' ${podname}-${c1name} ${podname}-${c2name} ${podname}-${c3name}
is "${lines[0]}" "false" "ReadonlyRootfs - container 1, post-run"
is "${lines[1]}" "true" "ReadonlyRootfs - container 2, post-run"
is "${lines[2]}" "true" "ReadonlyRootfs - container 3, post-run"
# Clean up
run_podman kube down - < $YAML
run_podman 1 container exists ${podname}-${c1name}
run_podman 1 container exists ${podname}-${c2name}
run_podman 1 container exists ${podname}-${c3name}
}
# bats test_tags=ci:parallel
@test "podman play with user from image" {
imgname="i-$(safename)"
_write_test_yaml command=id image=$imgname
cat > $PODMAN_TMPDIR/Containerfile << _EOF
from $IMAGE
USER bin
_EOF
# Unset the PATH during build and make sure that all default env variables
# are correctly set for the created container.
# --layers=false needed to work around buildah#5674 parallel flake
run_podman build --layers=false --unsetenv PATH -t $imgname $PODMAN_TMPDIR
run_podman image inspect $imgname --format "{{.Config.Env}}"
is "$output" "\[\]" "image does not set PATH - env is empty"
run_podman play kube --start=false $TESTYAML
run_podman inspect --format "{{ .Config.User }}" $PODCTRNAME
is "$output" bin "expect container within pod to run as the bin user"
run_podman inspect --format "{{ .Config.Env }}" $PODCTRNAME
is "$output" ".*PATH=/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin.*" "expect PATH to be set"
is "$output" ".*container=podman.*" "expect container to be set"
run_podman pod rm -t 0 -f $PODNAME
run_podman rmi -f $imgname
}
# CANNOT BE PARALLELIZED (YET): buildah#5674, parallel builds fail
# ...workaround is --layers=false, but there's no way to do that in kube
@test "podman play --build --context-dir" {
skip_if_remote "--build is not supported in context remote"
imgname="userimage-$(safename)"
mkdir -p $PODMAN_TMPDIR/$imgname
cat > $PODMAN_TMPDIR/$imgname/Containerfile << _EOF
from $IMAGE
USER bin
_EOF
_write_test_yaml command=id image=quay.io/libpod/$imgname
run_podman 125 play kube --build --start=false $TESTYAML
assert "$output" =~ "initializing source docker://quay.io/libpod/$imgname:latest: reading manifest latest in "
run_podman play kube --replace --context-dir=$PODMAN_TMPDIR --build --start=false $TESTYAML
run_podman inspect --format "{{ .Config.User }}" $PODCTRNAME
is "$output" bin "expect container within pod to run as the bin user"
run_podman pod rm -t 0 -f $PODNAME
run_podman rmi -f $imgname
cd $PODMAN_TMPDIR
run_podman play kube --replace --build --start=false $TESTYAML
run_podman inspect --format "{{ .Config.User }}" $PODCTRNAME
is "$output" bin "expect container within pod to run as the bin user"
run_podman pod rm -t 0 -f $PODNAME
run_podman rmi -f $imgname
}
# Occasionally a remnant storage container is left behind which causes
# podman play kube --replace to fail. This tests created a conflicting
# storage container name using buildah to make sure --replace, still
# functions properly by removing the storage container.
# bats test_tags=ci:parallel
@test "podman kube play --replace external storage" {
_write_test_yaml command="top"
run_podman play kube $TESTYAML
# Pod container exists
run_podman ps -a --format '=={{.Names}}=={{.Status}}=='
assert "$output" =~ "==${PODCTRNAME}==Up " "Pod container created"
# Force removal of container, and confirm that it no longer exists
run_podman rm --force -t0 $PODCTRNAME
assert "$output" = "$PODCTRNAME" "Confirmation that container was removed"
run_podman ps -a --external --format '=={{.Names}}=={{.Status}}=='
assert "$output" !~ "$PODCTRNAME" "Pod container gone"
# Create external container using buildah with same name
buildah from --name $PODCTRNAME $IMAGE
# Confirm that we now have an external (Storage) container by that name
run_podman ps -a --external --format '=={{.Names}}=={{.Status}}=='
assert "$output" =~ "==${PODCTRNAME}==Storage==" "Storage (external) container created"
# --replace deletes the buildah container and replace it with new one.
# Prior to #20025, this would fail with "container name is in use"
run_podman play kube --replace $TESTYAML
run_podman pod rm -t 0 -f $PODNAME
}
# bats test_tags=ci:parallel
@test "podman kube --annotation" {
_write_test_yaml command=/home/podman/pause
RANDOMSTRING=$(random_string 15)
ANNOTATION_WITH_COMMA="comma,$(random_string 5)"
run_podman kube play --annotation "name=$RANDOMSTRING" \
--annotation "anno=$ANNOTATION_WITH_COMMA" $TESTYAML
run_podman inspect --format "{{ .Config.Annotations }}" $PODCTRNAME
is "$output" ".*name:$RANDOMSTRING" "Annotation should be added to pod"
is "$output" ".*anno:$ANNOTATION_WITH_COMMA" "Annotation with comma should be added to pod"
# invalid annotation
run_podman 125 kube play --annotation "val" $TESTYAML
assert "$output" == "Error: annotation \"val\" must include an '=' sign" "invalid annotation error"
run_podman pod rm -t 0 -f $PODNAME
}
# bats test_tags=ci:parallel
@test "podman play Yaml deprecated --no-trunc annotation" {
skip "FIXME: I can't figure out what this test is supposed to do"
RANDOMSTRING=$(random_string 65)
_write_test_yaml "annotations=test: ${RANDOMSTRING}" command=id
run_podman play kube --no-trunc - < $TESTYAML
}
# bats test_tags=ci:parallel
@test "podman kube play - default log driver" {
_write_test_yaml command=top
# Get the default log driver
run_podman info --format "{{.Host.LogDriver}}"
default_driver=$output
# Make sure that the default log driver is used
run_podman kube play $TESTYAML
run_podman inspect --format "{{.HostConfig.LogConfig.Type}}" $PODCTRNAME
is "$output" "$default_driver" "play kube uses default log driver"
run_podman kube down $TESTYAML
run_podman 125 inspect $PODCTRNAME
is "$output" ".*Error: no such object: \"$PODCTRNAME\""
}
# bats test_tags=ci:parallel
@test "podman kube play - URL" {
_write_test_yaml command=top
echo READY > $PODMAN_TMPDIR/ready
HOST_PORT=$(random_free_port)
SERVER=http://127.0.0.1:$HOST_PORT
serverctr="yamlserver-$(safename)"
run_podman run -d --name $serverctr -p "$HOST_PORT:80" \
-v $TESTYAML:/var/www/testpod.yaml:Z \
-v $PODMAN_TMPDIR/ready:/var/www/ready:Z \
-w /var/www \
$IMAGE /bin/busybox-extras httpd -f -p 80
wait_for_port 127.0.0.1 $HOST_PORT
wait_for_command_output "curl -s -S $SERVER/ready" "READY"
run_podman kube play $SERVER/testpod.yaml
run_podman inspect $PODCTRNAME --format "{{.State.Running}}"
is "$output" "true"
run_podman kube down $SERVER/testpod.yaml
run_podman 125 inspect $PODCTRNAME
is "$output" ".*Error: no such object: \"$PODCTRNAME\""
run_podman rm -f -t0 $serverctr
}
# bats test_tags=ci:parallel
@test "podman play with init container" {
_write_test_yaml command=
cat >>$TESTYAML <<EOF
- command:
- ls
- /dev/shm/test1
image: $IMAGE
name: ${CTRNAME}-test
initContainers:
- command:
- touch
- /dev/shm/test1
image: $IMAGE
name: ${CTRNAME}-init
EOF
run_podman kube play $TESTYAML
assert "$output" !~ "level=" "init containers should not generate logrus.Error"
run_podman inspect --format "{{.State.ExitCode}}" ${PODNAME}-${CTRNAME}-test
is "$output" "0" "init container should have created /dev/shm/test1"
run_podman kube down $TESTYAML
}
# bats test_tags=ci:parallel
@test "podman kube play - hostport" {
HOST_PORT=$(random_free_port)
_write_test_yaml
cat >>$TESTYAML <<EOF
- name: ${CTRNAME}-server
image: $IMAGE
ports:
- name: hostp
hostPort: $HOST_PORT
EOF
run_podman kube play $TESTYAML
run_podman pod inspect $PODNAME --format "{{.InfraConfig.PortBindings}}"
assert "$output" = "map[$HOST_PORT/tcp:[{0.0.0.0 $HOST_PORT}]]"
run_podman kube down $TESTYAML
}
# bats test_tags=ci:parallel
@test "podman kube play - multi-pod YAML" {
skip_if_remote "service containers only work locally"
skip_if_journald_unavailable
# Create the YAMl file, with two pods, each with one container
podnamebase="p-$(safename)"
ctrnamebase="c-$(safename)"
for n in 1 2;do
_write_test_yaml labels="app: $podnamebase-$n" name="$podnamebase-$n" ctrname="$ctrnamebase-$n" command=top
# Separator between two yaml halves
if [[ $n = 1 ]]; then
echo "---" >>$TESTYAML
fi
done
# Run `play kube` in the background as it will wait for the service
# container to exit.
timeout --foreground -v --kill=10 60 \
$PODMAN play kube --service-container=true --log-driver journald $TESTYAML &>/dev/null &
# The name of the service container is predictable: the first 12 characters
# of the hash of the YAML file followed by the "-service" suffix
yaml_sha=$(sha256sum $TESTYAML)
service_container="${yaml_sha:0:12}-service"
# Wait for the containers to be running
container_1="${podnamebase}-1-${ctrnamebase}-1"
container_2="${podnamebase}-2-${ctrnamebase}-2"
containers_running=
for i in $(seq 1 20); do
run_podman "?" container wait $container_1 $container_2 $service_container --condition="running"
if [[ $status == 0 ]]; then
containers_running=1
break
fi
sleep 0.5
# Just for debugging
run_podman ps -a
done
if [[ -z "$containers_running" ]]; then
die "container $container_1, $container_2 and/or $service_container did not start"
fi
# Stop the pods, make sure that no ugly error logs show up and that the
# service container will implicitly get stopped as well
run_podman pod stop ${podnamebase}-1 ${podnamebase}-2
assert "$output" !~ "Stopping"
_ensure_container_running $service_container false
run_podman kube down $TESTYAML
}
# bats test_tags=ci:parallel
@test "podman kube generate filetype" {
YAML=$PODMAN_TMPDIR/test.yml
podname="p-$(safename)"
ctrname="c-$(safename)"
volname="v-$(safename)"
run_podman create --pod new:$podname \
--security-opt label=level:s0:c1,c2 \
--security-opt label=filetype:usr_t \
-v $volname:/myvol --name $ctrname $IMAGE true
run_podman kube generate $podname -f $YAML
run cat $YAML
is "$output" ".*filetype: usr_t" "Generated YAML file should contain filetype usr_t"
run_podman pod rm --force $podname
run_podman volume rm -t -1 $volname --force
run_podman kube play $YAML
if selinux_enabled; then
run_podman inspect ${podname}-${ctrname} --format "{{ .MountLabel }}"
is "$output" "system_u:object_r:usr_t:s0:c1,c2" "Generated container should use filetype usr_t"
run_podman volume inspect $volname --format '{{ .Mountpoint }}'
path=${output}
run ls -Zd $path
is "$output" "system_u:object_r:usr_t:s0 $path" "volume should be labeled with usr_t type"
fi
run_podman kube down $YAML
run_podman volume rm $volname --force
}
# kube play --wait=true, where we clear up the created containers, pods, and volumes when a kill or sigterm is triggered
# bats test_tags=ci:parallel
@test "podman kube play --wait with siginterrupt" {
podname="p-$(safename)"
ctrname="c-$(safename)"
fname="$PODMAN_TMPDIR/play_kube_wait_$(random_string 6).yaml"
echo "
apiVersion: v1
kind: Pod
metadata:
labels:
app: test
name: $podname
spec:
restartPolicy: Never
containers:
- name: $ctrname
image: $IMAGE
command:
- top
- -b
" > $fname
# Run in background, then wait for pod to start running.
# This guarantees that when we send the signal (below) we do so
# on a running container; signaling during initialization
# results in undefined behavior.
logfile=$PODMAN_TMPDIR/kube-play.log
$PODMAN kube play --wait $fname &> $logfile &
local kidpid=$!
for try in {1..10}; do
run_podman '?' container inspect --format '{{.State.Running}}' "$podname-$ctrname"
if [[ $status -eq 0 ]] && [[ "$output" = "true" ]]; then
break
fi
sleep 1
done
wait_for_output "Mem:" "$podname-$ctrname"
# Send SIGINT to container, and see how long it takes to exit.
local t0=$SECONDS
kill -2 $kidpid
wait $kidpid
local t1=$SECONDS
local delta_t=$((t1 - t0))
# Expectation (in seconds) of when we should time out. When running
# parallel, allow longer time due to system load
local expect=4
if [[ -n "$PARALLEL_JOBSLOT" ]]; then
expect=$((expect + 4))
fi
assert $delta_t -le $expect \
"podman kube play did not get killed within $expect seconds"
# Make sure we actually got SIGTERM and podman printed its message.
assert "$(< $logfile)" =~ "Cleaning up containers, pods, and volumes" \
"kube play printed sigterm message"
# there should be no containers running or created
run_podman ps -a --noheading
assert "$output" !~ "$(safename)" "All containers created by this test should be gone"
# ...nor pods
run_podman pod ps --noheading
assert "$output" !~ "$(safename)" "All pods created by this test should be gone"
}
# bats test_tags=ci:parallel
@test "podman kube play --wait - wait for pod to exit" {
podname="p-$(safename)"
ctrname="c-$(safename)"
fname="$PODMAN_TMPDIR/play_kube_wait_$(random_string 6).yaml"
echo "
apiVersion: v1
kind: Pod
metadata:
labels:
app: test
name: $podname
spec:
restartPolicy: Never
containers:
- name: $ctrname
image: $IMAGE
command:
- echo
- hello
" > $fname
run_podman kube play --wait $fname
# debug to see what container is being left behind after the cleanup
# there should be no containers running or created
run_podman ps -a --noheading
assert "$output" !~ "$(safename)" "No containers created by this test"
run_podman pod ps
assert "$output" !~ "$(safename)" "No pods created by this test"
}
# bats test_tags=ci:parallel
@test "podman kube play with configmaps" {
foovalue="foo-$(safename)"
barvalue="bar-$(safename)"
configmap_file=${PODMAN_TMPDIR}/play_kube_configmap_configmaps$(random_string 6),withcomma.yaml
echo "
---
apiVersion: v1
kind: ConfigMap
metadata:
name: foo
data:
value: $foovalue
---
apiVersion: v1
kind: ConfigMap
metadata:
name: bar
data:
value: $barvalue
" > $configmap_file
podname="p-$(safename)"
ctrname="c-$(safename)"
pod_file=${PODMAN_TMPDIR}/play_kube_configmap_pod$(random_string 6).yaml
echo "
apiVersion: v1
kind: Pod
metadata:
labels:
app: test
name: $podname
spec:
restartPolicy: Never
containers:
- name: $ctrname
image: $IMAGE
env:
- name: FOO
valueFrom:
configMapKeyRef:
name: foo
key: value
- name: BAR
valueFrom:
configMapKeyRef:
name: bar
key: value
command:
- /bin/sh
args:
- -c
- "echo \$FOO:\$BAR"
" > $pod_file
run_podman kube play --configmap=$configmap_file $pod_file
run_podman wait $podname-$ctrname
# systemd logs are unreliable; we may need to retry a few times
# https://github.com/systemd/systemd/issues/28650
local retries=10
while [[ $retries -gt 0 ]]; do
run_podman logs $podname-$ctrname
test -n "$output" && break
sleep 0.1
retries=$((retries - 1))
done
assert "$retries" -gt 0 "Timed out waiting for podman logs"
assert "$output" = "$foovalue:$barvalue" "output from podman logs"
run_podman kube down $pod_file
}
# bats test_tags=ci:parallel
@test "podman kube with --authfile=/tmp/bogus" {
_write_test_yaml
bogus=$PODMAN_TMPDIR/bogus-authfile
run_podman 125 kube play --authfile=$bogus - < $TESTYAML
is "$output" "Error: credential file is not accessible: faccessat $bogus: no such file or directory" \
"$command should fail with not such file"
}
# bats test_tags=ci:parallel
@test "podman kube play with umask from containers.conf" {
skip_if_remote "remote does not support CONTAINERS_CONF*"
YAML=$PODMAN_TMPDIR/test.yaml
containersConf=$PODMAN_TMPDIR/containers.conf
touch $containersConf
cat >$containersConf <<EOF
[containers]
umask = "0472"
EOF
ctr="c-$(safename)"
ctrInPod="${ctr}-pod-${ctr}"
run_podman create --restart never --name $ctr $IMAGE sh -c "touch /umask-test;stat -c '%a' /umask-test"
run_podman kube generate -f $YAML $ctr
CONTAINERS_CONF_OVERRIDE="$containersConf" run_podman kube play $YAML
run_podman container inspect --format '{{ .Config.Umask }}' $ctrInPod
is "${output}" "0472"
# Confirm that umask actually takes effect. Might take a second or so.
local retries=10
while [[ $retries -gt 0 ]]; do
run_podman logs $ctrInPod
test -n "$output" && break
sleep 0.1
retries=$((retries - 1))
done
assert "$retries" -gt 0 "Timed out waiting for container output"
assert "$output" = "204" "stat() on created file"
run_podman kube down $YAML
run_podman rm $ctr
}
# bats test_tags=ci:parallel
@test "podman kube generate tmpfs on /tmp" {
_write_test_yaml command=/home/podman/pause
run_podman kube play $TESTYAML
run_podman exec $PODCTRNAME sh -c "mount | grep /tmp"
assert "$output" !~ "noexec" "mounts on /tmp should not be noexec"
run_podman kube down $TESTYAML
}
# bats test_tags=ci:parallel
@test "podman kube play - pull policy" {
skip_if_remote "pull debug logs only work locally"
_write_test_yaml command=true
# Exploit a debug message to make sure the expected pull policy is used
run_podman --debug kube play $TESTYAML
assert "$output" =~ "Pulling image $IMAGE \(policy\: missing\)" "default pull policy is missing"
run_podman kube down $TESTYAML
local_image="localhost/i-$(safename):latest"
run_podman tag $IMAGE $local_image
rm $TESTYAML
_write_test_yaml command=true image=$local_image
run_podman --debug kube play $TESTYAML
assert "$output" =~ "Pulling image $local_image \(policy\: newer\)" "pull policy is set to newhen pulling latest tag"
run_podman kube down $TESTYAML
run_podman rmi $local_image
}
# bats test_tags=ci:parallel
@test "podman kube play healthcheck should wait initialDelaySeconds before updating status" {
for want in healthy unhealthy; do
podname="liveness-exec-$(safename)-$want"
ctrname="liveness-ctr-$(safename)-$want"
fname="$PODMAN_TMPDIR/play_kube_${want}_$(random_string 6).yaml"
cat <<EOF >$fname
apiVersion: v1
kind: Pod
metadata:
labels:
name: $podname
spec:
containers:
- name: $ctrname
image: $IMAGE
args:
- /bin/sh
- -c
- touch /tmp/healthy && sleep 100
livenessProbe:
exec:
# /tmp/healthy will exist in healthy container
# /tmp/unhealthy will never exist, and will thus
# cause healthcheck failure
command:
- cat
- /tmp/$want
initialDelaySeconds: 3
failureThreshold: 1
periodSeconds: 1
EOF
run_podman kube play $fname
ctrName="$podname-$ctrname"
# Collect status every half-second until it goes into the desired state.
local i=1
local full_log=""
while [[ $i -lt 15 ]]; do
run_podman inspect $ctrName --format "$i-{{.State.Health.Status}}"
full_log+=" $output"
if [[ "$output" =~ "-$want" ]]; then
break
fi
sleep 0.5
i=$((i+1))
done
assert "$full_log" =~ "-$want\$" \
"Container got to '$want'"
assert "$full_log" =~ "-starting.*-$want" \
"Container went from starting to $want"
if [[ $want == "healthy" ]]; then
dontwant="unhealthy"
else
dontwant="healthy"
fi
assert "$full_log" !~ "-$dontwant" \
"Container never goes $dontwant"
# GAH! Save ten seconds, but in a horrible way.
# - 'kube down' does not have a -t0 option.
# - Using 'top' in the container, instead of 'sleep 100', results
# in very weird failures. Seriously weird.
# - 'stop -t0', every once in a while on parallel runs on my
# laptop (never yet in CI), barfs with 'container is running or
# paused, refusing to clean up, container state improper'
# Here's hoping that this will silence the flakes.
run_podman '?' stop -t0 $ctrName
run_podman kube down $fname
done
}
# CANNOT BE PARALLELIZED (YET): buildah#5674, parallel builds fail
# ...workaround is --layers=false, but there's no way to do that in kube
@test "podman play --build private registry" {
skip_if_remote "--build is not supported in context remote"
local registry=localhost:${PODMAN_LOGIN_REGISTRY_PORT}
local from_image=$registry/quadlet_image_test-$(safename):$(random_string)
local authfile=$PODMAN_TMPDIR/authfile.json
userimage=userimage-$(safename)
mkdir -p $PODMAN_TMPDIR/$userimage
cat > $PODMAN_TMPDIR/$userimage/Containerfile << _EOF
from $from_image
USER bin
_EOF
# Start the registry and populate the authfile that we can use for the test.
start_registry
run_podman login --authfile=$authfile \
--tls-verify=false \
--username ${PODMAN_LOGIN_USER} \
--password ${PODMAN_LOGIN_PASS} \
$registry
# Push the test image to the registry
run_podman image tag $IMAGE $from_image
run_podman image push --tls-verify=false --authfile=$authfile $from_image
# Remove the local image to make sure it will be pulled again
run_podman image rm --ignore $from_image
# The error below assumes unqualified-search registries exist, however the default
# distro config may not set some and thus resulting in a different error message.
# We could try to match a third or or simply force a know static config to trigger
# the right error.
local CONTAINERS_REGISTRIES_CONF="$PODMAN_TMPDIR/registries.conf"
echo 'unqualified-search-registries = ["quay.io"]' > "$CONTAINERS_REGISTRIES_CONF"
export CONTAINERS_REGISTRIES_CONF
_write_test_yaml command=id image=$userimage
run_podman 125 play kube --build --start=false $TESTYAML
assert "$output" "=~" \
"Error: short-name resolution enforced but cannot prompt without a TTY|Resolving \"$userimage\" using unqualified-search registries" \
"The error message does match any of the expected ones"
run_podman play kube --replace --context-dir=$PODMAN_TMPDIR --tls-verify=false --authfile=$authfile --build --start=false $TESTYAML
run_podman inspect --format "{{ .Config.User }}" $PODCTRNAME
is "$output" bin "expect container within pod to run as the bin user"
run_podman kube down $TESTYAML
run_podman rmi -f $userimage $from_image
}
# bats test_tags=ci:parallel
@test "podman play with image volume (automount annotation and OCI VolumeSource)" {
imgname1="automount-img1-$(safename)"
imgname2="automount-img2-$(safename)"
podname="p-$(safename)"
ctrname="c-$(safename)"
ctrname_not_mounted="c-not-mounted-$(safename)"
cat >$PODMAN_TMPDIR/Containerfile1 <<EOF
FROM $IMAGE
RUN mkdir /test1 /test_same && \
touch /test1/a /test1/b /test1/c && \
echo "I am from test1 image" > /test_same/hello_world
VOLUME /test1
VOLUME /test_same
EOF
cat >$PODMAN_TMPDIR/Containerfile2 <<EOF
FROM $IMAGE
RUN mkdir /test2 /test_same && \
touch /test2/asdf /test2/ejgre /test2/lteghe && \
echo "I am from test2 image" > /test_same/hello_world
VOLUME /test2
VOLUME /test_same
EOF
# --layers=false needed to work around buildah#5674 parallel flake
run_podman build -t $imgname1 --layers=false -f $PODMAN_TMPDIR/Containerfile1
run_podman build -t $imgname2 --layers=false -f $PODMAN_TMPDIR/Containerfile2
_write_test_yaml command=top name=$podname ctrname=$ctrname
run_podman kube play --annotation "io.podman.annotations.kube.image.volumes.mount/$ctrname=$imgname1" $TESTYAML
run_podman run --rm $imgname1 ls -x /test1
assert "$output" = "a b c" "ls /test1 on image"
run_out_test1="$output"
run_podman exec $podname-$ctrname ls -x /test1
assert "$output" = "$run_out_test1" "matching ls run/exec volume path test1"
run_podman run --rm $imgname1 cat /test_same/hello_world
assert "$output" = "I am from test1 image" "cat /test_same/hello_world on image"
run_out_hello_world="$output"
run_podman exec $podname-$ctrname cat /test_same/hello_world
assert "$output" = "$run_out_hello_world" "matching cat /test_same/hello_world volume path test_same"
run_podman kube down $TESTYAML
# Testing the first technique to mount an OCI image: through a Pod annotation
fname="/$PODMAN_TMPDIR/play_kube_wait_$(random_string 6).yaml"
cat >$fname <<EOF
apiVersion: v1
kind: Pod
metadata:
labels:
app: test
name: $podname
annotations:
io.podman.annotations.kube.image.volumes.mount/$ctrname: $imgname1;$imgname2
spec:
restartPolicy: Never
containers:
- name: $ctrname
image: $IMAGE
command:
- top
- name: $ctrname_not_mounted
image: $IMAGE
command:
- top
EOF
run_podman kube play $fname
run_podman exec "$podname-$ctrname" ls -x /test1
assert "a b c" "ls /test1 inside container"
run_podman exec "$podname-$ctrname" ls -x /test2
assert "asdf ejgre lteghe" "ls /test2 inside container"
run_podman exec "$podname-$ctrname" cat /test_same/hello_world
assert "I am from test2 image" "cat /test_same/hello_world inside container"
run_podman 1 exec "$podname-$ctrname" touch /test1/readonly
assert "$output" =~ "Read-only file system" "image mounted as readonly"
run_podman exec "$podname-$ctrname_not_mounted" ls /
assert "$output" !~ "test" "No volume should be mounted in no-mount container"
run_podman kube down $fname
# Testing the second technique to mount an OCI image: using image volume
fname="/$PODMAN_TMPDIR/play_kube_wait_$(random_string 6).yaml"
cat >$fname <<EOF
apiVersion: v1
kind: Pod
metadata:
labels:
app: test
name: $podname
spec:
restartPolicy: Never
containers:
- name: $ctrname
image: $IMAGE
command:
- top
volumeMounts:
- name: volume1
mountPath: /image1
- name: volume2
mountPath: /image2
- name: $ctrname_not_mounted
image: $IMAGE
command:
- top
volumes:
- name: volume1
image:
reference: $imgname1
- name: volume2
image:
reference: $imgname2
EOF
run_podman kube play $fname
run_podman exec "$podname-$ctrname" ls -x /image1/test1
assert "a b c" "ls /test1 inside container"
run_podman exec "$podname-$ctrname" ls -x /image2/test2
assert "asdf ejgre lteghe" "ls /test2 inside container"
run_podman 1 exec "$podname-$ctrname" touch /image1/test1/readonly
assert "$output" =~ "Read-only file system" "image mounted as readonly"
run_podman exec "$podname-$ctrname_not_mounted" ls /
assert "$output" !~ "image" "No volume should be mounted in no-mount container"
run_podman kube down $fname
run_podman rmi $imgname1 $imgname2
}
# bats test_tags=ci:parallel
@test "podman play with image volume pull policies" {
podname="p-$(safename)"
ctrname="c-$(safename)"
volimg_local="localhost/i-$(safename):latest" # only exists locally
volimg_remote=${PODMAN_NONLOCAL_IMAGE_FQN} # only exists remotely
volimg_both="quay.io/libpod/alpine:latest" # exists both remotely and locally
localfile="localfile"
# Pull $volimg_both and commit a local modification. As a result
# the image exists both locally and remotely but the two versions
# are slightly different.
run_podman pull $volimg_both
run_podman run --name $ctrname $volimg_both sh -c "touch /$localfile"
run_podman commit $ctrname $volimg_both
run_podman rm $ctrname
# Tag $volimg_both as $volimg_local
run_podman tag $volimg_both $volimg_local
# Check that $volimg_both and $volimg_local exists locally
run_podman image exists $volimg_both
run_podman image exists $volimg_local
# Check that $volimg_remote doesn't exist locally
run_podman 1 image exists $volimg_remote
# The test scenarios:
# We verify the return code of kube play for different
# combinations of image locations (remote/local/both)
# and pull policies (IfNotPresent/Always/Never).
# When the image exists both locally and remotely the
# tests verify that the correct version of image is
# mounted (0 for the local version, 1 for the remote
# version).
# When running the tests with $volimg_both, the test
# with policy "Always" should be run as the last one,
# as it pulls the remote image and overwrites the local
# one.
tests="
$volimg_local | IfNotPresent | 0 |
$volimg_local | Never | 0 |
$volimg_local | Always | 125 |
$volimg_remote | IfNotPresent | 0 |
$volimg_remote | Never | 125 |
$volimg_remote | Always | 0 |
$volimg_both | IfNotPresent | 0 | 0
$volimg_both | Never | 0 | 0
$volimg_both | Always | 0 | 1
"
while read volimg policy playrc islocal; do
fname="/$PODMAN_TMPDIR/play_kube_volimg_${policy}.yaml"
cat >$fname <<EOF
apiVersion: v1
kind: Pod
metadata:
labels:
app: test
name: $podname
spec:
restartPolicy: Never
containers:
- name: $ctrname
image: $IMAGE
command:
- top
volumeMounts:
- name: volume
mountPath: /image
volumes:
- name: volume
image:
reference: $volimg
pullPolicy: $policy
EOF
run_podman $playrc kube play $fname
if [[ "$islocal" != "''" ]]; then
run_podman $islocal exec "$podname-$ctrname" ls /image/$localfile
fi
run_podman kube down $fname
# If the remote-only image was pulled, remove it
run_podman rmi -f $volimg_remote
done < <(parse_table "$tests")
run_podman rmi $volimg_local $volimg_both
}
# CANNOT BE PARALLELIZED: userns=auto, rootless, => not enough unused IDs in user namespace
@test "podman kube restore user namespace" {
if ! is_rootless; then
grep -E -q "^containers:" /etc/subuid || skip "no IDs allocated for user 'containers'"
fi
podname="p-userns-$(safename)"
run_podman pod create --userns auto --name $podname
run_podman create --pod $podname $IMAGE true
run_podman pod inspect --format {{.InfraContainerID}} $podname
infraID="$output"
run_podman inspect --format '{{index .Config.Annotations "io.podman.annotations.userns"}}' $infraID
assert "$output" == "auto" "user namespace should be kept"
YAML=$PODMAN_TMPDIR/test.yml
# Make sure the same setting is restored if the pod is recreated from the yaml
run_podman kube generate $podname -f $YAML
cat $YAML
run_podman kube play --replace $YAML
run_podman pod inspect --format {{.InfraContainerID}} $podname
infraID="$output"
run_podman inspect --format '{{index .Config.Annotations "io.podman.annotations.userns"}}' $infraID
assert "$output" == "auto" "user namespace should be kept"
run_podman pod rm -f $podname
}