Merge pull request #9229 from ldoktor/ocp-ci

ocp.ci: Various fixes and improvements to the OCP pipeline
This commit is contained in:
Wainer Moschetta 2024-03-19 11:13:01 -03:00 committed by GitHub
commit e6501aa4ad
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
5 changed files with 167 additions and 11 deletions

55
ci/openshift-ci/cleanup.sh Executable file
View File

@ -0,0 +1,55 @@
#!/bin/bash
#
# Copyright (c) 2024 Red Hat, Inc.
#
# SPDX-License-Identifier: Apache-2.0
#
# This script tries to removes most of the resources added by `test.sh` script
# from the cluster.
scripts_dir=$(dirname $0)
deployments_dir=${scripts_dir}/cluster/deployments
configs_dir=${scripts_dir}/configs
source ${scripts_dir}/lib.sh
# Set to 'yes' if you want to configure SELinux to permissive on the cluster
# workers.
#
SELINUX_PERMISSIVE=${SELINUX_PERMISSIVE:-no}
# Enable workaround for OCP 4.13 https://github.com/kata-containers/kata-containers/pull/9206
#
WORKAROUND_9206_CRIO=${WORKAROUND_9206_CRIO:-no}
# Ignore errors as we want best-effort-approach here
trap - ERR
# Delete potential smoke-test resources
oc delete -f "${scripts_dir}/smoke/service.yaml"
oc delete -f "${scripts_dir}/smoke/service_kubernetes.yaml"
oc delete -f "${scripts_dir}/smoke/http-server.yaml"
# Delete test.sh resources
oc delete -f "${deployments_dir}/relabel_selinux.yaml"
if [[ "$WORKAROUND_9206_CRIO" == "yes" ]]; then
oc delete -f "${deployments_dir}/workaround-9206-crio-ds.yaml"
oc delete -f "${deployments_dir}/workaround-9206-crio.yaml"
fi
[ ${SELINUX_PERMISSIVE} == "yes" ] && oc delete -f "${deployments_dir}/machineconfig_selinux.yaml.in"
# Delete kata-containers
pushd "$katacontainers_repo_dir/tools/packaging/kata-deploy"
oc delete -f kata-deploy/base/kata-deploy.yaml
oc -n kube-system wait --timeout=10m --for=delete -l name=kata-deploy pod
oc apply -f kata-cleanup/base/kata-cleanup.yaml
echo "Wait for all related pods to be gone"
( repeats=1; for i in $(seq 1 600); do
oc get pods -l name="kubelet-kata-cleanup" --no-headers=true -n kube-system 2>&1 | grep "No resources found" -q && ((repeats++)) || repeats=1
[ "$repeats" -gt 5 ] && echo kata-cleanup finished && break
sleep 1
done) || { echo "There are still some kata-cleanup related pods after 600 iterations"; oc get all -n kube-system; exit -1; }
oc delete -f kata-cleanup/base/kata-cleanup.yaml
oc delete -f kata-rbac/base/kata-rbac.yaml
oc delete -f runtimeclasses/kata-runtimeClasses.yaml

View File

@ -23,6 +23,7 @@ spec:
set -e;
echo Starting the relabel;
nsenter --target 1 --mount bash -xc '
command -v semanage &>/dev/null || { echo Does not look like a SELINUX cluster, skipping; exit 0; };
for ENTRY in \
\"/(.*/)?opt/kata/bin(/.*)?\" \
\"/(.*/)?opt/kata/runtime-rs/bin(/.*)?\" \
@ -31,7 +32,7 @@ spec:
\"/(.*/)?opt/kata/share/tdvf(/.*)?\" \
\"/(.*/)?opt/kata/libexec(/.*)?\";
do
semanage fcontext -a -t qemu_exec_t \"$ENTRY\" || { echo \"Error in semanage command\"; exit 1; }
semanage fcontext -a -t qemu_exec_t \"$ENTRY\" || semanage fcontext -m -t qemu_exec_t \"$ENTRY\" || { echo \"Error in semanage command\"; exit 1; }
done;
restorecon -v -R /opt/kata || { echo \"Error in restorecon command\"; exit 1; }
';

View File

@ -0,0 +1,28 @@
---
apiVersion: apps/v1
kind: DaemonSet
metadata:
name: workaround-9206-crio-ds
spec:
selector:
matchLabels:
app: workaround-9206-crio-ds
template:
metadata:
labels:
app: workaround-9206-crio-ds
spec:
containers:
- name: workaround-9206-crio-ds
image: alpine
volumeMounts:
- name: host-dir
mountPath: /tmp/config
securityContext:
runAsUser: 0
privileged: true
command: ["/bin/sh", "-c", "while [ ! -f '/tmp/config/10-workaround-9206-crio' ]; do sleep 1; done; echo 'Config file present'; sleep infinity"]
volumes:
- name: host-dir
hostPath:
path: /etc/crio/crio.conf.d/

View File

@ -0,0 +1,18 @@
---
apiVersion: machineconfiguration.openshift.io/v1
kind: MachineConfig
metadata:
labels:
machineconfiguration.openshift.io/role: worker
name: 10-workaround-9206-crio
spec:
config:
ignition:
version: 2.2.0
storage:
files:
- contents:
source: data:text/plain;charset=utf-8;base64,W2NyaW9dCnN0b3JhZ2Vfb3B0aW9uID0gWwoJIm92ZXJsYXkuc2tpcF9tb3VudF9ob21lPXRydWUiLApdCg==
filesystem: root
mode: 0644
path: /etc/crio/crio.conf.d/10-workaround-9206-crio

View File

@ -27,16 +27,21 @@ KATA_WITH_SYSTEM_QEMU=${KATA_WITH_SYSTEM_QEMU:-no}
#
KATA_WITH_HOST_KERNEL=${KATA_WITH_HOST_KERNEL:-no}
# kata-deploy image to be used to deploy the kata (by default use CI image
# that is built for each pull request)
#
KATA_DEPLOY_IMAGE=${KATA_DEPLOY_IMAGE:-quay.io/kata-containers/kata-deploy-ci:kata-containers-latest}
# Enable workaround for OCP 4.13 https://github.com/kata-containers/kata-containers/pull/9206
#
WORKAROUND_9206_CRIO=${WORKAROUND_9206_CRIO:-no}
# Leverage kata-deploy to install Kata Containers in the cluster.
#
apply_kata_deploy() {
local deploy_file="tools/packaging/kata-deploy/kata-deploy/base/kata-deploy.yaml"
local old_img="quay.io/kata-containers/kata-deploy:latest"
# Use the kata-deploy CI image which is built for each pull request merged
local new_img="quay.io/kata-containers/kata-deploy-ci:kata-containers-latest"
pushd "$katacontainers_repo_dir"
sed -i "s#${old_img}#${new_img}#" "$deploy_file"
sed -ri "s#(\s+image:) .*#\1 ${KATA_DEPLOY_IMAGE}#" "$deploy_file"
info "Applying kata-deploy"
oc apply -f tools/packaging/kata-deploy/kata-rbac/base/kata-rbac.yaml
@ -91,7 +96,7 @@ wait_for_reboot() {
}
wait_mcp_update() {
local delta="${1:-900}"
local delta="${1:-1200}"
local sleep_time=30
# The machineconfigpool is fine when all the workers updated and are ready,
# and none are degraded.
@ -147,6 +152,51 @@ debug_pod() {
oc logs "$pod"
}
# Wait for all pods of the app label to contain expected message
#
# Params:
# $1 - app labela
# $2 - expected pods count (>=1)
# $3 - message to be present in the logs
# $4 - timeout (60)
# $5 - namespace (the current one)
wait_for_app_pods_message() {
local app="$1"
local pod_count="$2"
local message="$3"
local timeout="$4"
local namespace="$5"
[ -z "$pod_count" ] && pod_count=1
[ -z "$timeout" ] && timeout=60
[ -n "$namespace" ] && namespace=" -n $namespace "
local pod
local pods
local i
SECONDS=0
while :; do
pods=($(oc get pods -l app="$app" --no-headers=true $namespace | awk '{print $1}'))
[ "${#pods}" -ge "$pod_count" ] && break
if [ "$SECONDS" -gt "$timeout" ]; then
echo "Unable to find ${pod_count} pods for '-l app=\"$app\"' in ${SECONDS}s (${pods[@]})"
return -1
fi
done
for pod in "${pods[@]}"; do
while :; do
local log=$(oc logs $namespace "$pod")
echo "$log" | grep "$message" -q && echo "Found $(echo "$log" | grep "$message") in $pod's log ($SECONDS)" && break;
if [ "$SECONDS" -gt "$timeout" ]; then
echo -n "Message '$message' not present in '${pod}' pod of the '-l app=\"$app\"' "
echo "pods after ${SECONDS}s (${pods[@]})"
echo "Pod $pod's output so far:"
echo "$log"
return -1
fi
sleep 1;
done
done
}
oc config set-context --current --namespace=default
worker_nodes=$(oc get nodes | awk '{if ($3 == "worker") { print $1 } }')
@ -182,10 +232,14 @@ if [ ${SELINUX_PERMISSIVE} == "yes" ]; then
wait_for_reboot
fi
if [[ "$WORKAROUND_9206_CRIO" == "yes" ]]; then
info "Applying workaround to enable skip_mount_home in crio on OCP 4.13"
oc apply -f "${deployments_dir}/workaround-9206-crio.yaml"
oc apply -f "${deployments_dir}/workaround-9206-crio-ds.yaml"
wait_for_app_pods_message workaround-9206-crio-ds "$num_nodes" "Config file present" 1200 || echo "Failed to apply the workaround, proceeding anyway..."
fi
# FIXME: Remove when https://github.com/kata-containers/kata-containers/pull/8417 is resolved
# Selinux context is currently not handled by kata-deploy
oc apply -f ${deployments_dir}/relabel_selinux.yaml
( for I in $(seq 30); do
sleep 10
oc logs -n kube-system ds/relabel-selinux-daemonset | grep "NSENTER_FINISHED_WITH:" && exit
done ) || { echo "Selinux relabel failed, check the logs"; exit -1; }
wait_for_app_pods_message restorecon "$num_nodes" "NSENTER_FINISHED_WITH:" 120 "kube-system" || echo "Failed to treat selinux, proceeding anyway..."