"ec2:Describe*"
"ec2:StopInstances"
"ec2:ReplaceRoute"

OSの起動時間 >= 強制停止スクリプトのタイムアウト + ハートビートタイムアウト + ハートビートインターバル

#! /bin/sh
#***********************************************
#* forcestop.sh *
#***********************************************
# AWS CLIの絶対パス
AWS_CLI=”/usr/local/bin/aws” ★ AWS CLIのパス

# 1ノード目の設定情報
SERVER1_NAME="server01" ★ サーバー1のホスト名
SERVER1_INSTANCE="i-11111111111111111" ★ サーバー1のインスタンスID

# 2ノード目の設定情報
SERVER2_NAME="server02" ★ サーバー2のホスト名
SERVER2_INSTANCE="i-22222222222222222" ★ サーバー2のインスタンスID

# インスタンス停止の最大試行回数
STOP_LOOP_MAX=2 ★ 停止処理の最大試行回数

# インスタンス状態確認の最大試行回数
CHECK_LOOP_MAX=120 ★ 確認処理の最大試行回数

echo "START"
echo "DOWN SERVER NAME : ${CLP_SERVER_DOWN}"
echo "LOCAL SERVER NAME: ${CLP_SERVER_LOCAL}"
if [ "${CLP_SERVER_DOWN}" = "${SERVER1_NAME}" ]; then
  INSTANCE="${SERVER1_INSTANCE}"
elif [ "${CLP_SERVER_DOWN}" = "${SERVER2_NAME}" ]; then
  INSTANCE=${SERVER2_INSTANCE}
else
  echo "DOWN SERVER is not found."
  echo "EXIT 1" exit 1
fi

# シャットダウン要求
STOP_LOOP_COUNT=0

while [ ${STOP_LOOP_COUNT} -lt ${STOP_LOOP_MAX} ]
do
  ${AWS_CLI} ec2 stop-instances --instance-ids ${INSTANCE} --force
  ret=$?
  if [ ${ret} -eq 0 ]; then
    echo "succeeded to stop instance. (${INSTANCE})"
    break
  fi
  echo "failed to stop instance. (${INSTANCE}, ret=${ret})"

  sleep 1
  let STOP_LOOP_COUNT=${STOP_LOOP_COUNT}+1
done

if [ ${ret} -ne 0 ]; then
  echo "EXIT ${ret}"
  exit ${ret}
fi

# サーバーダウン確認
# サーバーの停止を確認できなくても、CHECK_LOOP_MAXを超えた場合は停止が
# 完了したものとして待ち合わせ処理を終了する
CHECK_LOOP_COUNT=0
while [ ${CHECK_LOOP_COUNT} -lt ${CHECK_LOOP_MAX} ]
do
  ${AWS_CLI} ec2 describe-instances --instance-ids ${INSTANCE} --filters "Name=instance-state-name,Values=stopped" | grep ${INSTANCE}
  if [ $? -eq 0 ]; then
    echo "${INSTANCE} has been stopped."
    break
  fi

  sleep 1
  let CHECK_LOOP_COUNT=${CHECK_LOOP_COUNT}+1
done

echo "EXIT 0"
exit 0

[[root@server01 ~]# clpstat
 ========================  CLUSTER STATUS  ===========================
  Cluster : cluster
  <server>
   *server01 ........: Online     ←server01は起動中
      lankhb1        : Normal           Kernel Mode LAN Heartbeat
      pingnp1        : Normal           ping resolution
    server02 ........: Offline    ←server02は停止中
      lankhb1        : Unknown          Kernel Mode LAN Heartbeat
      pingnp1        : Unknown          ping resolution
  <group>
    failover ........: Online
      current        : server01   ←server01でフェールオーバーグループが起動中
      awsvip         : Online
      md             : Online
  <monitor>
    awsvipw1         : Normal
    mdnw1            : Error
    mdw1             : Error
    userw            : Normal
 =====================================================================

[root@server01 ~]#

[[root@server02 ~]# clpstat
 ========================  CLUSTER STATUS  ===========================
  Cluster : cluster
  <server>
    server01 ........: Offline    ←server01は停止中
      lankhb1        : Unknown          Kernel Mode LAN Heartbeat
      pingnp1        : Unknown          ping resolution
   *server02 ........: Online     ←server02は起動中
      lankhb1        : Normal           Kernel Mode LAN Heartbeat
      pingnp1        : Normal           ping resolution
  <group>
    failover ........: Online
      current        : server02   ←server02でフェールオーバーグループが起動中
      awsvip         : Online
      md             : Online
  <monitor>
    awsvipw1         : Normal
    mdnw1            : Error
    mdw1             : Error
    userw            : Normal
 =====================================================================

[root@server02 ~]#

Jun 14 05:05:24 server02 clusterpro: [E] <type: rm><event: 9> Detected an error in monitoring mdnw1. (31 : The network is disconnected.)
Jun 14 05:05:57 server02 clusterpro: [I] <type: nm><event: 4> Resource lankhb1 of server server01 has stopped.
Jun 14 05:05:57 server02 clusterpro: [I] <type: nm><event: 2> Server server01 has been stopped.            ★現用系サーバーのダウンを検出
Jun 14 05:05:58 server02 clusterpro: [I] <type: rc><event: 411> Script for forced stop has started.        ★強制停止スクリプトを実行開始
Jun 14 05:07:50 server02 clusterpro: [I] <type: rc><event: 412> Script for forced stop has completed.      ★強制停止スクリプトの実行完了
Jun 14 05:07:50 server02 clusterpro: [I] <type: rc><event: 60> Failover group failover has started.        ★フェールオーバーグループの起動を開始
Jun 14 05:07:50 server02 clusterpro: [I] <type: rc><event: 10> Activating group failover has started.
Jun 14 05:07:50 server02 clusterpro: [I] <type: rc><event: 30> Activating awsvip resource has started.

[root@server02 ~]# aws ec2 describe-instances --instance-ids i-11111111111111111 --query 'Reservations[0].Instances[0].[InstanceId, State.Name]'
 [
    " i-11111111111111111",
    "stopped"
]
[root@server02 ~]#