installers/install_k3s_cluster.sh

   1 #!/bin/bash
   2 #
   3 #   Licensed under the Apache License, Version 2.0 (the "License");
   4 #   you may not use this file except in compliance with the License.
   5 #   You may obtain a copy of the License at
   6 #
   7 #       http://www.apache.org/licenses/LICENSE-2.0
   8 #
   9 #   Unless required by applicable law or agreed to in writing, software
  10 #   distributed under the License is distributed on an "AS IS" BASIS,
  11 #   WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
  12 #   See the License for the specific language governing permissions and
  13 #   limitations under the License.
  14 #
  15
  16 set +eux
  17
  18 # K3s releases: https://github.com/k3s-io/k3s/releases/
  19 K8S_VERSION="v1.29.3+k3s1"
  20
  21 # installs k3s
  22 function install_k3s() {
  23     [ -z "${DEBUG_INSTALL}" ] || DEBUG beginning of function
  24     export INSTALL_K3S_EXEC="--disable traefik"
  25     curl -sfL https://get.k3s.io | INSTALL_K3S_VERSION=${K8S_VERSION} sh -s -
  26     sudo chmod 644 /etc/rancher/k3s/k3s.yaml
  27     [ -z "${DEBUG_INSTALL}" ] || DEBUG end of function
  28 }
  29
  30 # updates service nodeport range
  31 function update_service_nodeport_range() {
  32     [ -z "${DEBUG_INSTALL}" ] || DEBUG beginning of function
  33     sudo k3s server --kube-apiserver-arg=service-node-port-range=80-32767
  34     [ -z "${DEBUG_INSTALL}" ] || DEBUG end of function
  35 }
  36
  37 # checks cluster readiness
  38 function check_for_readiness() {
  39     [ -z "${DEBUG_INSTALL}" ] || DEBUG beginning of function
  40     # Check for Ready node, takes ~30 seconds
  41     echo "Waiting for K8s nodes to be ready"
  42     local time_for_failure=60    # seconds broken
  43     local sampling_period=5     # seconds
  44     local counter=0
  45     local cluster_ready=""
  46     while (( counter < time_for_failure ))
  47     do
  48         kubectl get nodes |grep master |grep -v none | grep Ready
  49         if [ $? -eq 0 ] ; then
  50             echo "K8s cluster is ready"
  51             cluster_ready="y"
  52             break
  53         else
  54             echo "K8s cluster is not ready yet"
  55             counter=$((counter + sampling_period))
  56             sleep ${sampling_period}
  57         fi
  58     done
  59     [ -n "$cluster_ready" ] || FATAL_TRACK k8scluster "K3s cluster nodes not ready after $time_for_failure seconds."
  60
  61     echo "Waiting for pods to be ready"
  62     local time_for_readiness=20   # seconds ready
  63     local time_for_failure=100    # seconds broken
  64
  65     # Equivalent number of samples
  66     oks_threshold=$((time_for_readiness/${sampling_period}))     # No. ok samples to declare the system ready
  67     failures_threshold=$((time_for_failure/${sampling_period}))  # No. nok samples to declare the system broken
  68     failures_in_a_row=0
  69     oks_in_a_row=0
  70     ####################################################################################
  71     # Loop to check system readiness
  72     ####################################################################################
  73     K3S_NAMESPACE=kube-system
  74     while [[ (${failures_in_a_row} -lt ${failures_threshold}) && (${oks_in_a_row} -lt ${oks_threshold}) ]]
  75     do
  76         # State of pods rather than completed jobs
  77         K3S_PODS_STATE=$(kubectl get pod -n ${K3S_NAMESPACE} --no-headers |grep -v Completed 2>&1)
  78         K3S_PODS_READY=$(echo "${K3S_PODS_STATE}" | awk '$2=="1/1" || $2=="2/2" {printf ("%s\t%s\t\n", $1, $2)}')
  79         K3S_PODS_NOT_READY=$(echo "${K3S_PODS_STATE}" | awk '$2!="1/1" && $2!="2/2" {printf ("%s\t%s\t\n", $1, $2)}')
  80         COUNT_K3S_PODS_READY=$(echo "${K3S_PODS_READY}"| grep -v -e '^$' | wc -l)
  81         COUNT_K3S_PODS_NOT_READY=$(echo "${K3S_PODS_NOT_READY}" | grep -v -e '^$' | wc -l)
  82
  83         # OK sample
  84         if [[ ${COUNT_K3S_PODS_NOT_READY} -eq 0 ]]
  85         then
  86             ((++oks_in_a_row))
  87             failures_in_a_row=0
  88             echo -ne ===\> Successful checks: "${oks_in_a_row}"/${oks_threshold}\\r
  89         # NOK sample
  90         else
  91             ((++failures_in_a_row))
  92             oks_in_a_row=0
  93             echo
  94             echo Bootstraping... "${failures_in_a_row}" checks of ${failures_threshold}
  95
  96             # Reports failed pods in K3S
  97             if [[ "${COUNT_K3S_PODS_NOT_READY}" -ne 0 ]]
  98             then
  99                 echo "K3S kube-system: Waiting for ${COUNT_K3S_PODS_NOT_READY} of $((${COUNT_K3S_PODS_NOT_READY}+${COUNT_K3S_PODS_READY})) pods to be ready:"
 100                 echo "${K3S_PODS_NOT_READY}"
 101                 echo
 102             fi
 103         fi
 104
 105         #------------ NEXT SAMPLE
 106         sleep ${sampling_period}
 107     done
 108
 109     ####################################################################################
 110     # OUTCOME
 111     ####################################################################################
 112     if [[ (${failures_in_a_row} -ge ${failures_threshold}) ]]
 113     then
 114         echo
 115         FATAL_TRACK k8scluster "K8S CLUSTER IS BROKEN"
 116     else
 117         echo
 118         echo "K8S CLUSTER IS READY"
 119     fi
 120     [ -z "${DEBUG_INSTALL}" ] || DEBUG end of function
 121 }
 122
 123 # Initializes kubeconfig file
 124 function save_kubeconfig() {
 125     [ -z "${DEBUG_INSTALL}" ] || DEBUG beginning of function
 126     KUBEDIR="${HOME}/.kube"
 127     KUBEFILE="$KUBEDIR/config"
 128     mkdir -p "${KUBEDIR}"
 129     K3S_KUBECONFIG="/etc/rancher/k3s/k3s.yaml"
 130     sudo cp "${K3S_KUBECONFIG}" "${KUBEFILE}"
 131     sudo chown $(id -u):$(id -g) "${KUBEFILE}"
 132     sed -i "s#server: https://127.0.0.1#server: https://${DEFAULT_IP}#g" "${KUBEFILE}"
 133     chmod 700 "${KUBEFILE}"
 134     echo
 135     echo "Credentials saved at ${KUBEFILE}"
 136     echo
 137     [ -z "${DEBUG_INSTALL}" ] || DEBUG end of function
 138 }
 139
 140 # main
 141 DEFAULT_IP="127.0.0.1"
 142 OSM_DEVOPS="/usr/share/osm-devops"
 143 DEBUG_INSTALL=""
 144
 145 while getopts ":D:i:-: " o; do
 146     case "${o}" in
 147         i)
 148             DEFAULT_IP="${OPTARG}"
 149             ;;
 150         D)
 151             OSM_DEVOPS="${OPTARG}"
 152             ;;
 153         -)
 154             [ "${OPTARG}" == "debug" ] && DEBUG_INSTALL="y" && continue
 155             echo -e "Invalid option: '--$OPTARG'\n" >&2
 156             exit 1
 157             ;;
 158         :)
 159             echo "Option -$OPTARG requires an argument" >&2
 160             exit 1
 161             ;;
 162         \?)
 163             echo -e "Invalid option: '-$OPTARG'\n" >&2
 164             exit 1
 165             ;;
 166         *)
 167             exit 1
 168             ;;
 169     esac
 170 done
 171
 172 source $OSM_DEVOPS/common/logging
 173 source $OSM_DEVOPS/common/track
 174
 175 echo "DEBUG_INSTALL=${DEBUG_INSTALL}"
 176 echo "DEFAULT_IP=${DEFAULT_IP}"
 177 echo "OSM_DEVOPS=${OSM_DEVOPS}"
 178 echo "HOME=$HOME"
 179
 180 install_k3s
 181 track k8scluster k3s_install_ok
 182 check_for_readiness
 183 track k8scluster k3s_node_ready_ok
 184 # update_service_nodeport_range
 185 # check_for_readiness
 186 # track k8scluster k3s_update_nodeport_range_ok
 187 save_kubeconfig
 188 track k8scluster k3s_creds_ok