-#!/bin/bash -ex
+#!/bin/bash -e
+##############################################################################
+# Copyright (c) 2016 NEC Corporation and others.
#
-# Copyright 2016 NEC Corporation.
-#
-# Licensed under the Apache License, Version 2.0 (the "License"); you may
-# not use this file except in compliance with the License. You may obtain
-# a copy of the License at
-#
-# http://www.apache.org/licenses/LICENSE-2.0
-#
-# Unless required by applicable law or agreed to in writing, software
-# distributed under the License is distributed on an "AS IS" BASIS, WITHOUT
-# WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the
-# License for the specific language governing permissions and limitations
-# under the License.
+# All rights reserved. This program and the accompanying materials
+# are made available under the terms of the Apache License, Version 2.0
+# which accompanies this distribution, and is available at
+# http://www.apache.org/licenses/LICENSE-2.0
+##############################################################################
-#branch=$(git rev-parse --abbrev-ref HEAD)
-BRANCH=master
+[[ "${CI_DEBUG:-true}" == "true" ]] && set -x
IMAGE_URL=https://launchpad.net/cirros/trunk/0.3.0/+download/cirros-0.3.0-x86_64-disk.img
IMAGE_NAME=cirros
IMAGE_FORMAT=qcow2
VM_NAME=doctor_vm1
VM_FLAVOR=m1.tiny
-COMPUTE_HOST='s142'
ALARM_NAME=doctor_alarm1
INSPECTOR_PORT=12345
CONSUMER_PORT=12346
+DOCTOR_USER=doctor
+DOCTOR_PW=doctor
+DOCTOR_PROJECT=doctor
+#TODO: change back to `_member_` when JIRA DOCTOR-55 is done
+DOCTOR_ROLE=admin
+
+SUPPORTED_INSTALLER_TYPES="apex local"
+INSTALLER_TYPE=${INSTALLER_TYPE:-apex}
+INSTALLER_IP=${INSTALLER_IP:-none}
+COMPUTE_HOST=${COMPUTE_HOST:-overcloud-novacompute-0}
+COMPUTE_IP=${COMPUTE_IP:-none}
+COMPUTE_USER=${COMPUTE_USER:-heat-admin}
+ssh_opts="-o UserKnownHostsFile=/dev/null -o StrictHostKeyChecking=no"
+
+if [[ ! "$SUPPORTED_INSTALLER_TYPES" =~ "$INSTALLER_TYPE" ]] ; then
+ echo "ERROR: INSTALLER_TYPE=$INSTALLER_TYPE is not supported."
+ exit 1
+fi
+
+prepare_compute_ssh() {
+ ssh_opts_cpu="$ssh_opts"
+
+ if [[ "$INSTALLER_TYPE" == "apex" ]] ; then
+ if [[ "$INSTALLER_IP" == "none" ]] ; then
+ instack_mac=$(sudo virsh domiflist instack | awk '/default/{print $5}')
+ INSTALLER_IP=$(/usr/sbin/arp -e | grep ${instack_mac} | awk '{print $1}')
+ fi
+
+ if [[ "$COMPUTE_IP" == "none" ]] ; then
+ COMPUTE_IP=$(sudo ssh $ssh_opts $INSTALLER_IP \
+ "source stackrc; \
+ nova show $COMPUTE_HOST \
+ | awk '/ ctlplane network /{print \$5}'")
+ fi
+
+ # get ssh key from installer node
+ sudo scp $ssh_opts root@"$INSTALLER_IP":/home/stack/.ssh/id_rsa instack_key
+ sudo chown $(whoami):$(whoami) instack_key
+ chmod 400 instack_key
+ ssh_opts_cpu+=" -i instack_key"
+ elif [[ "$INSTALLER_TYPE" == "local" ]] ; then
+ if [[ "$COMPUTE_IP" == "none" ]] ; then
+ COMPUTE_IP=$(getent hosts "$COMPUTE_HOST" | awk '{ print $1 }')
+ if [[ -z "$COMPUTE_IP" ]]; then
+ echo "ERROR: Could not resolve $COMPUTE_HOST. Either manually set COMPUTE_IP or enable DNS resolution."
+ exit 1
+ fi
+ fi
+
+ echo "INSTALLER_TYPE set to 'local'. Assuming SSH keys already exchanged with $COMPUTE_HOST"
+ fi
+ # verify connectivity to target compute host
+ ping -c 1 "$COMPUTE_IP"
+}
download_image() {
[ -e "$IMAGE_FILE" ] && return 0
--file "$IMAGE_FILE"
}
+create_test_user() {
+ openstack user list | grep -q " $DOCTOR_USER " || {
+ openstack user create "$DOCTOR_USER" --password "$DOCTOR_PW"
+ }
+ openstack project list | grep -q " $DOCTOR_PROJECT " || {
+ openstack project create "$DOCTOR_PROJECT"
+ }
+ openstack user role list "$DOCTOR_USER" --project "$DOCTOR_PROJECT" \
+ | grep -q " $DOCTOR_ROLE " || {
+ openstack role add "$DOCTOR_ROLE" --user "$DOCTOR_USER" \
+ --project "$DOCTOR_PROJECT"
+ }
+}
+
+change_to_doctor_user() {
+ export OS_USERNAME="$DOCTOR_USER"
+ export OS_PASSWORD="$DOCTOR_PW"
+ export OS_PROJECT_NAME="$DOCTOR_PROJECT"
+ export OS_TENANT_NAME="$DOCTOR_PROJECT"
+}
+
boot_vm() {
- nova list | grep -q " $VM_NAME " && return 0
- nova boot --flavor "$VM_FLAVOR" \
- --image "$IMAGE_NAME" \
- "$VM_NAME"
- sleep 1
+ (
+ # test VM done with test user, so can test non-admin
+ change_to_doctor_user
+ nova list | grep -q " $VM_NAME " && return 0
+ nova boot --flavor "$VM_FLAVOR" \
+ --image "$IMAGE_NAME" \
+ "$VM_NAME"
+ sleep 1
+ )
+
}
create_alarm() {
- ceilometer alarm-list | grep -q " $ALARM_NAME " && return 0
- vm_id=$(nova list | grep " $VM_NAME " | awk '{print $2}')
- ceilometer alarm-event-create --name "$ALARM_NAME" \
- --alarm-action "http://localhost:$CONSUMER_PORT/failure" \
- --description "VM failure" \
- --enabled True \
- --repeat-actions False \
- --severity "moderate" \
- --event-type compute.instance.update \
- -q "traits.state=string::error; traits.instance_id=string::$vm_id"
+ (
+ # get vm_id as test user
+ change_to_doctor_user
+ ceilometer alarm-list | grep -q " $ALARM_NAME " && return 0
+ vm_id=$(nova list | grep " $VM_NAME " | awk '{print $2}')
+ ceilometer alarm-event-create --name "$ALARM_NAME" \
+ --alarm-action "http://localhost:$CONSUMER_PORT/failure" \
+ --description "VM failure" \
+ --enabled True \
+ --repeat-actions False \
+ --severity "moderate" \
+ --event-type compute.instance.update \
+ -q "traits.state=string::error; traits.instance_id=string::$vm_id"
+ )
}
+
start_monitor() {
pgrep -f "python monitor.py" && return 0
- sudo python monitor.py "$COMPUTE_HOST" "http://127.0.0.1:$INSPECTOR_PORT/events" > monitor.log 2>&1 &
- MONITOR_PID=$!
+ sudo python monitor.py "$COMPUTE_HOST" "$COMPUTE_IP" \
+ "http://127.0.0.1:$INSPECTOR_PORT/events" > monitor.log 2>&1 &
}
stop_monitor() {
wait_for_vm_launch() {
echo "waiting for vm launch..."
- while true
- do
- state=$(nova list | grep " $VM_NAME " | awk '{print $6}')
- [[ "$state" == "ACTIVE" ]] && return 0
- sleep 1
- done
+
+ (
+ # get VM state as test user
+ change_to_doctor_user
+
+ count=0
+ while [[ ${count} -lt 60 ]]
+ do
+ state=$(nova list | grep " $VM_NAME " | awk '{print $6}')
+ [[ "$state" == "ACTIVE" ]] && return 0
+ [[ "$state" == "ERROR" ]] && echo "vm state is ERROR" && exit 1
+ count=$(($count+1))
+ sleep 1
+ done
+ echo "ERROR: time out while waiting for vm launch"
+ exit 1
+ )
}
inject_failure() {
- #FIXME
- echo ssh $COMPUTE_HOST "ip link set eno1 down"
+ echo "disabling network of compute host [$COMPUTE_HOST] for 3 mins..."
+ cat > disable_network.sh << 'END_TXT'
+#!/bin/bash -x
+dev=$(sudo ip route | awk '/^default/{print $5}')
+sleep 1
+sudo ip link set $dev down
+sleep 180
+sudo ip link set $dev up
+sleep 1
+END_TXT
+ chmod +x disable_network.sh
+ scp $ssh_opts_cpu disable_network.sh "$COMPUTE_USER@$COMPUTE_IP:"
+ ssh $ssh_opts_cpu "$COMPUTE_USER@$COMPUTE_IP" 'nohup ./disable_network.sh > disable_network.log 2>&1 &'
}
calculate_notification_time() {
- detect=$(grep "doctor monitor detected at" monitor.log | awk '{print $5}')
+ detected=$(grep "doctor monitor detected at" monitor.log | awk '{print $5}')
notified=$(grep "doctor consumer notified at" consumer.log | awk '{print $5}')
- duration=$(echo "$notified $detect" | awk '{print $1 - $2 }')
- echo "$notified $detect" | \
- awk '{d = $1 - $2; if (d < 1 ) print d " OK"; else print d " NG"}'
+ echo "$notified $detected" | \
+ awk '{d = $1 - $2; if (d < 1 && d > 0) print d " OK"; else print d " NG"}'
+}
+
+check_host_status() {
+ expect_state=$1
+ (
+ change_to_doctor_user
+
+ host_status_line=$(nova show $VM_NAME | grep "host_status")
+ if [[ $? -ne 0 ]] ; then
+ echo "ERROR: host_status not configured for owner in Nova policy.json"
+ exit 1
+ fi
+
+ host_status=$(echo $host_status_line | awk '{print $4}')
+ if [ -z "$host_status" ] ; then
+ echo "ERROR: host_status not reported by: nova show $VM_NAME"
+ exit 1
+ elif [[ "$host_status" != "$expect_state" ]] ; then
+ echo "ERROR: host_status:$host_status not equal to expect_state: $expect_state"
+ exit 1
+ else
+ echo "$VM_NAME showing host_status: $host_status"
+ fi
+ )
}
-# TODO(r-mibu): Make sure env params are set properly for OpenStack clients
+cleanup() {
+ set +e
+ echo "cleanup..."
+ stop_monitor
+ stop_inspector
+ stop_consumer
+ python ./nova_force_down.py "$COMPUTE_HOST" --unset
+ sleep 1
+ (
+ change_to_doctor_user
+ nova list | grep -q " $VM_NAME " && nova delete "$VM_NAME"
+ sleep 1
+ alarm_id=$(ceilometer alarm-list | grep " $ALARM_NAME " | awk '{print $2}')
+ sleep 1
+ [ -n "$alarm_id" ] && ceilometer alarm-delete "$alarm_id"
+ sleep 1
+ )
+ image_id=$(glance image-list | grep " $IMAGE_NAME " | awk '{print $2}')
+ sleep 1
+ [ -n "$image_id" ] && glance image-delete "$image_id"
+ openstack role remove "$DOCTOR_ROLE" --user "$DOCTOR_USER" \
+ --project "$DOCTOR_PROJECT"
+ openstack project delete "$DOCTOR_PROJECT"
+ openstack user delete "$DOCTOR_USER"
+
+ echo "waiting disabled compute host back to be enabled..."
+ sleep 180
+ check_host_status "UP"
+ ssh $ssh_opts_cpu "$COMPUTE_USER@$COMPUTE_IP" \
+ "[ -e disable_network.log ] && cat disable_network.log"
+}
+
+
+echo "Note: doctor/tests/run.sh has been executed."
+
+prepare_compute_ssh
+
+trap cleanup EXIT
+
+echo "preparing VM image..."
download_image
register_image
+echo "starting doctor sample components..."
start_monitor
start_inspector
start_consumer
+echo "creating test user..."
+create_test_user
+
+echo "creating VM and alarm..."
boot_vm
create_alarm
wait_for_vm_launch
sleep 60
+echo "injecting host failure..."
inject_failure
sleep 10
+check_host_status "DOWN"
calculate_notification_time
echo "done"
Code Review / doctor.git / blobdiff