2 ##############################################################################
3 # Copyright (c) 2016 NEC Corporation and others.
5 # All rights reserved. This program and the accompanying materials
6 # are made available under the terms of the Apache License, Version 2.0
7 # which accompanies this distribution, and is available at
8 # http://www.apache.org/licenses/LICENSE-2.0
9 ##############################################################################
11 IMAGE_URL=https://launchpad.net/cirros/trunk/0.3.0/+download/cirros-0.3.0-x86_64-disk.img
13 IMAGE_FILE="${IMAGE_NAME}.img"
17 ALARM_NAME=doctor_alarm1
21 INSTALLER_TYPE=${INSTALLER_TYPE:-apex}
22 INSTALLER_IP=${INSTALLER_IP:-none}
23 COMPUTE_HOST=${COMPUTE_HOST:-overcloud-novacompute-0}
24 COMPUTE_IP=${COMPUTE_IP:-none}
25 ssh_opts="-o UserKnownHostsFile=/dev/null -o StrictHostKeyChecking=no"
27 if [[ "$INSTALLER_TYPE" != "apex" ]] ; then
28 echo "ERROR: INSTALLER_TYPE=$INSTALLER_TYPE is not supported."
32 if [[ "$INSTALLER_IP" == "none" ]] ; then
33 instack_mac=$(sudo virsh domiflist instack | awk '/default/{print $5}')
34 INSTALLER_IP=$(/usr/sbin/arp -e | grep ${instack_mac} | awk '{print $1}')
37 if [[ "$COMPUTE_IP" == "none" ]] ; then
38 COMPUTE_IP=$(sudo ssh $ssh_opts $INSTALLER_IP \
40 nova show $COMPUTE_HOST \
41 | awk '/ ctlplane network /{print \$5}'")
44 prepare_compute_ssh() {
45 ping -c 1 "$COMPUTE_IP"
47 # get ssh key from installer node
48 sudo scp $ssh_opts /home/stack/.ssh/id_rsa instack_key
49 if [ ! -r instack_key ]; then
50 sudo chown $(whoami):$(whoami) instack_key
53 ssh_opts_cpu="$ssh_opts -i instack_key -l heat-admin"
57 [ -e "$IMAGE_FILE" ] && return 0
58 wget "$IMAGE_URL" -o "$IMAGE_FILE"
62 glance image-list | grep -q " $IMAGE_NAME " && return 0
63 glance image-create --name "$IMAGE_NAME" \
65 --disk-format "$IMAGE_FORMAT" \
66 --container-format bare \
71 nova list | grep -q " $VM_NAME " && return 0
72 nova boot --flavor "$VM_FLAVOR" \
73 --image "$IMAGE_NAME" \
79 ceilometer alarm-list | grep -q " $ALARM_NAME " && return 0
80 vm_id=$(nova list | grep " $VM_NAME " | awk '{print $2}')
81 ceilometer alarm-event-create --name "$ALARM_NAME" \
82 --alarm-action "http://localhost:$CONSUMER_PORT/failure" \
83 --description "VM failure" \
85 --repeat-actions False \
86 --severity "moderate" \
87 --event-type compute.instance.update \
88 -q "traits.state=string::error; traits.instance_id=string::$vm_id"
92 pgrep -f "python monitor.py" && return 0
93 sudo python monitor.py "$COMPUTE_HOST" "$COMPUTE_IP" \
94 "http://127.0.0.1:$INSPECTOR_PORT/events" > monitor.log 2>&1 &
98 pgrep -f "python monitor.py" || return 0
99 sudo kill $(pgrep -f "python monitor.py")
104 pgrep -f "python inspector.py" && return 0
105 python inspector.py "$INSPECTOR_PORT" > inspector.log 2>&1 &
109 pgrep -f "python inspector.py" || return 0
110 kill $(pgrep -f "python inspector.py")
115 pgrep -f "python consumer.py" && return 0
116 python consumer.py "$CONSUMER_PORT" > consumer.log 2>&1 &
120 pgrep -f "python consumer.py" || return 0
121 kill $(pgrep -f "python consumer.py")
125 wait_for_vm_launch() {
126 echo "waiting for vm launch..."
129 state=$(nova list | grep " $VM_NAME " | awk '{print $6}')
130 [[ "$state" == "ACTIVE" ]] && return 0
136 echo "disabling network of comupte host [$COMPUTE_HOST] for 3 mins..."
137 cat > disable_network.sh << 'END_TXT'
139 dev=$(/usr/sbin/ip route | awk '/^default/{print $5}')
141 echo sudo ip link set $dev down
143 echo sudo ip link set $dev up
146 chmod +x disable_network.sh
147 scp $ssh_opts_cpu disable_network.sh "$COMPUTE_IP:"
148 ssh $ssh_opts_cpu "$COMPUTE_IP:" 'nohup ./disable_network.sh > disable_network.log 2>&1 &'
151 calculate_notification_time() {
152 detect=$(grep "doctor monitor detected at" monitor.log | awk '{print $5}')
153 notified=$(grep "doctor consumer notified at" consumer.log | awk '{print $5}')
154 duration=$(echo "$notified $detect" | awk '{print $1 - $2 }')
155 echo "$notified $detect" | \
156 awk '{d = $1 - $2; if (d < 1 ) print d " OK"; else print d " NG"}'
165 ssh $ssh_opts_cpu $COMPUTE_IP \
166 "[ -e disable_network.log ] && cat disable_network.log"
168 nova service-force-down --unset "$COMPUTE_HOST" nova-compute
170 nova delete "$VM_NAME"
172 alarm_id=$(ceilometer alarm-list | grep " $ALARM_NAME " | awk '{print $2}')
174 [ -n "$alarm_id" ] && ceilometer alarm-delete "$alarm_id"
176 image_id=$(glance image-list | grep " $IMAGE_NAME " | awk '{print $2}')
178 [ -n "$image_id" ] && glance image-delete "$image_id"
179 #TODO: add host status check via nova admin api
180 echo "waiting disabled compute host back to be enabled..."
185 echo "Note: doctor/tests/run.sh has been executed."
191 echo "preparing VM image..."
195 echo "starting doctor sample components..."
200 echo "creating VM and alarm..."
206 echo "injecting host failure..."
210 calculate_notification_time