#!/bin/bash # Activate robotframework virtualenv # ${ROBOT_VENV} comes from the integration-install-robotframework.sh # script. # shellcheck source=${ROBOT_VENV}/bin/activate disable=SC1091 source ${ROBOT_VENV}/bin/activate source /tmp/common-functions.sh ${BUNDLEFOLDER} # Ensure we fail the job if any steps fail. set -ex -o pipefail totaltmr=$(timer) get_os_deploy PYTHON="${ROBOT_VENV}/bin/python" SSH="ssh -t -t" ADMIN_PASSWORD="admin" OPENSTACK_MASTER_CLIENTS_VERSION="queens" #Size of the partition to /opt/stack in control and compute nodes TMPFS_SIZE=2G # TODO: remove this work to run changes.py if/when it's moved higher up to be visible at the Robot level printf "\nshowing recent changes that made it into the distribution used by this job:\n" $PYTHON -m pip install --upgrade urllib3 python ${WORKSPACE}/test/tools/distchanges/changes.py -d /tmp/distribution_folder \ -u ${ACTUAL_BUNDLE_URL} -b ${DISTROBRANCH} \ -r ssh://jenkins-${SILO}@git.opendaylight.org:29418 || true printf "\nshowing recent changes that made it into integration/test used by this job:\n" cd ${WORKSPACE}/test printf "Hash Author Date Commit Date Author Subject\n" printf "%s\n" "------- ------------------------------ ------------------------------ -------------------- -----------------------------" git --no-pager log --pretty=format:'%h %<(30)%ad %<(30)%cd %<(20,trunc)%an%d %s' -n20 printf "\n" cd - cat << EOF ################################################# ## Deploy Openstack 3-node ## ################################################# EOF # Catch command errors and collect logs. # This ensures logs are collected when script commands fail rather than simply exiting. function trap_handler() { local prog="$0" local lastline="$1" local lasterr="$2" echo "trap_handler: ${prog}: line ${lastline}: exit status of last command: ${lasterr}" echo "trap_handler: command: ${BASH_COMMAND}" exit 1 } # trap_handler() trap 'trap_handler ${LINENO} ${$?}' ERR print_job_parameters function create_etc_hosts() { NODE_IP=$1 CTRL_IP=$2 : > ${WORKSPACE}/hosts_file for iter in `seq 1 ${NUM_OPENSTACK_COMPUTE_NODES}`; do COMPUTE_IP=OPENSTACK_COMPUTE_NODE_${iter}_IP if [ "${!COMPUTE_IP}" == "${NODE_IP}" ]; then CONTROL_HNAME=$(${SSH} ${CTRL_IP} "hostname") echo "${CTRL_IP} ${CONTROL_HNAME}" >> ${WORKSPACE}/hosts_file else COMPUTE_HNAME=$(${SSH} ${!COMPUTE_IP} "hostname") echo "${!COMPUTE_IP} ${COMPUTE_HNAME}" >> ${WORKSPACE}/hosts_file fi done echo "Created the hosts file for ${NODE_IP}:" cat ${WORKSPACE}/hosts_file } # create_etc_hosts() #function to install Openstack Clients for Testing #This will pull the latest versions compatiable with the # openstack release function install_openstack_clients_in_robot_vm() { packages=("python-novaclient" "python-neutronclient" "python-openstackclient") local os_plugins os_plugins=$(csv2ssv "${ENABLE_OS_PLUGINS}") for plugin_name in $os_plugins; do if [ "$plugin_name" == "networking-sfc" ]; then packages+=("networking-sfc") fi done openstack_version=$(echo ${OPENSTACK_BRANCH} | cut -d/ -f2) #If the job tests "master", we will use the clients from previous released stable version to avoid failures if [ "${openstack_version}" == "master" ]; then openstack_version=${OPENSTACK_MASTER_CLIENTS_VERSION} fi for package in ${packages[*]}; do echo "Get the current support version of the package ${package}" wget https://raw.githubusercontent.com/openstack/requirements/stable/${openstack_version}/upper-constraints.txt -O /tmp/constraints.txt 2>/dev/null echo "$PYTHON -m pip install --upgrade --no-deps ${package} --no-cache-dir -c /tmp/constraints.txt" $PYTHON -m pip install --upgrade --no-deps ${package} --no-cache-dir -c /tmp/constraints.txt echo "$PYTHON -m pip install ${package} --no-cache-dir -c /tmp/constraints.txt" $PYTHON -m pip install ${package} --no-cache-dir -c /tmp/constraints.txt done if [ "${ENABLE_NETWORKING_L2GW}" == "yes" ]; then #networking-l2gw is not officially available in any release yet. Getting the latest stable version. $PYTHON -m pip install networking-l2gw==11.0.0 fi } #Function to install rdo release # This will help avoiding installing wrong version of packages which causes # functionality failures function install_rdo_release() { local ip=$1 case ${OPENSTACK_BRANCH} in *pike*) ${SSH} ${ip} "sudo yum install -y https://repos.fedorapeople.org/repos/openstack/openstack-pike/rdo-release-pike-1.noarch.rpm" ;; *queens*) ${SSH} ${ip} "sudo yum install -y https://repos.fedorapeople.org/repos/openstack/openstack-queens/rdo-release-queens-1.noarch.rpm" ;; master) ${SSH} ${ip} "sudo yum install -y https://repos.fedorapeople.org/repos/openstack/openstack-queens/rdo-release-queens-1.noarch.rpm" ;; esac } # Involves just setting up the shared directory function setup_live_migration_control() { local control_ip=$1 printf "${control_ip}:Setup directory Share with NFS" cat > ${WORKSPACE}/setup_live_migration_control.sh << EOF sudo mkdir --mode=777 /vm_instances sudo chown -R jenkins:jenkins /vm_instances sudo yum install -y nfs-utils printf "/vm_instances *(rw,no_root_squash)" | sudo tee -a /etc/exports sudo systemctl start rpcbind nfs-server sudo exportfs EOF scp ${WORKSPACE}/setup_live_migration_control.sh ${control_ip}:/tmp/setup_live_migration_control.sh ssh ${control_ip} "bash /tmp/setup_live_migration_control.sh" } # Involves mounting the share and configuring the libvirtd function setup_live_migration_compute() { local compute_ip=$1 local control_ip=$2 printf "${compute_ip}:Mount Shared directory from ${control_ip}" printf "${compute_ip}:Configure libvirt in listen mode" cat > ${WORKSPACE}/setup_live_migration_compute.sh << EOF sudo yum install -y libvirt libvirt-devel nfs-utils sudo crudini --verbose --set --inplace /etc/libvirt/libvirtd.conf '' listen_tls 0 sudo crudini --verbose --set --inplace /etc/libvirt/libvirtd.conf '' listen_tcp 1 sudo crudini --verbose --set --inplace /etc/libvirt/libvirtd.conf '' auth_tcp '"none"' sudo crudini --verbose --set --inplace /etc/sysconfig/libvirtd '' LIBVIRTD_ARGS '"--listen"' sudo mkdir --mode=777 -p /var/instances sudo chown -R jenkins:jenkins /var/instances sudo chmod o+x /var/instances sudo systemctl start rpcbind sudo mount -t nfs ${control_ip}:/vm_instances /var/instances sudo mount EOF scp ${WORKSPACE}/setup_live_migration_compute.sh ${compute_ip}:/tmp/setup_live_migration_compute.sh ssh ${compute_ip} "bash /tmp/setup_live_migration_compute.sh" } # Add enable_services and disable_services to the local.conf function add_os_services() { local core_services=$1 local enable_services=$2 local disable_services=$3 local local_conf_file_name=$4 local enable_network_services=$5 cat >> ${local_conf_file_name} << EOF enable_service $(csv2ssv "${core_services}") EOF if [ -n "${enable_services}" ]; then cat >> ${local_conf_file_name} << EOF enable_service $(csv2ssv "${enable_services}") EOF fi if [ -n "${disable_services}" ]; then cat >> ${local_conf_file_name} << EOF disable_service $(csv2ssv "${disable_services}") EOF fi if [ -n "${enable_network_services}" ]; then cat >> ${local_conf_file_name} << EOF enable_service $(csv2ssv "${enable_network_services}") EOF fi } function create_control_node_local_conf() { HOSTIP=$1 MGRIP=$2 ODL_OVS_MANAGERS="$3" local_conf_file_name=${WORKSPACE}/local.conf_control_${HOSTIP} cat > ${local_conf_file_name} << EOF [[local|localrc]] LOGFILE=stack.sh.log LOG_COLOR=False USE_SYSTEMD=True RECLONE=${RECLONE} # Increase the wait used by stack to poll for services SERVICE_TIMEOUT=120 disable_all_services EOF add_os_services "${CORE_OS_CONTROL_SERVICES}" "${ENABLE_OS_SERVICES}" "${DISABLE_OS_SERVICES}" "${local_conf_file_name}" "${ENABLE_OS_NETWORK_SERVICES}" cat >> ${local_conf_file_name} << EOF HOST_IP=${HOSTIP} SERVICE_HOST=\$HOST_IP Q_ML2_TENANT_NETWORK_TYPE=${TENANT_NETWORK_TYPE} NEUTRON_CREATE_INITIAL_NETWORKS=${CREATE_INITIAL_NETWORKS} ODL_MODE=manual ODL_MGR_IP=${MGRIP} ODL_PORT=${ODL_PORT} ODL_PORT_BINDING_CONTROLLER=${ODL_ML2_PORT_BINDING} ODL_OVS_MANAGERS=${ODL_OVS_MANAGERS} MYSQL_HOST=\$SERVICE_HOST RABBIT_HOST=\$SERVICE_HOST GLANCE_HOSTPORT=\$SERVICE_HOST:9292 KEYSTONE_AUTH_HOST=\$SERVICE_HOST KEYSTONE_SERVICE_HOST=\$SERVICE_HOST ADMIN_PASSWORD=${ADMIN_PASSWORD} DATABASE_PASSWORD=${ADMIN_PASSWORD} RABBIT_PASSWORD=${ADMIN_PASSWORD} SERVICE_TOKEN=${ADMIN_PASSWORD} SERVICE_PASSWORD=${ADMIN_PASSWORD} NEUTRON_LBAAS_SERVICE_PROVIDERV2=${LBAAS_SERVICE_PROVIDER} # Only relevant if neutron-lbaas plugin is enabled NEUTRON_SFC_DRIVERS=${ODL_SFC_DRIVER} # Only relevant if networking-sfc plugin is enabled NEUTRON_FLOWCLASSIFIER_DRIVERS=${ODL_SFC_DRIVER} # Only relevant if networking-sfc plugin is enabled ETCD_PORT=2379 PUBLIC_BRIDGE=${PUBLIC_BRIDGE} PUBLIC_PHYSICAL_NETWORK=${PUBLIC_PHYSICAL_NETWORK} ML2_VLAN_RANGES=${PUBLIC_PHYSICAL_NETWORK} ODL_PROVIDER_MAPPINGS=${ODL_PROVIDER_MAPPINGS} EOF if [ "${TENANT_NETWORK_TYPE}" == "local" ]; then cat >> ${local_conf_file_name} << EOF ENABLE_TENANT_TUNNELS=false EOF fi if [ "${ODL_ML2_DRIVER_VERSION}" == "v2" ]; then echo "ODL_V2DRIVER=True" >> ${local_conf_file_name} fi IFS=, for plugin_name in ${ENABLE_OS_PLUGINS}; do if [ "$plugin_name" == "networking-odl" ]; then ENABLE_PLUGIN_ARGS="${ODL_ML2_DRIVER_REPO} ${ODL_ML2_BRANCH}" elif [ "$plugin_name" == "kuryr-kubernetes" ]; then ENABLE_PLUGIN_ARGS="${DEVSTACK_KUBERNETES_PLUGIN_REPO} master" # note: kuryr-kubernetes only exists in master at the moment elif [ "$plugin_name" == "neutron-lbaas" ]; then ENABLE_PLUGIN_ARGS="${DEVSTACK_LBAAS_PLUGIN_REPO} ${OPENSTACK_BRANCH}" IS_LBAAS_PLUGIN_ENABLED="yes" elif [ "$plugin_name" == "networking-sfc" ]; then ENABLE_PLUGIN_ARGS="${DEVSTACK_NETWORKING_SFC_PLUGIN_REPO} ${OPENSTACK_BRANCH}" IS_SFC_PLUGIN_ENABLED="yes" else echo "Error: Invalid plugin $plugin_name, unsupported" continue fi cat >> ${local_conf_file_name} << EOF enable_plugin ${plugin_name} ${ENABLE_PLUGIN_ARGS} EOF done unset IFS if [ "${ENABLE_NETWORKING_L2GW}" == "yes" ]; then cat >> ${local_conf_file_name} << EOF enable_plugin networking-l2gw ${NETWORKING_L2GW_DRIVER} ${ODL_ML2_BRANCH} NETWORKING_L2GW_SERVICE_DRIVER=L2GW:OpenDaylight:networking_odl.l2gateway.driver_v2.OpenDaylightL2gwDriver:default EOF fi if [ "${ODL_ML2_DRIVER_VERSION}" == "v2" ]; then SERVICE_PLUGINS="odl-router_v2" else SERVICE_PLUGINS="odl-router" fi if [ "${ENABLE_NETWORKING_L2GW}" == "yes" ]; then SERVICE_PLUGINS+=", networking_l2gw.services.l2gateway.plugin.L2GatewayPlugin" fi if [ "${IS_LBAAS_PLUGIN_ENABLED}" == "yes" ]; then SERVICE_PLUGINS+=", lbaasv2" fi if [ "${IS_SFC_PLUGIN_ENABLED}" == "yes" ]; then SERVICE_PLUGINS+=", networking_sfc.services.flowclassifier.plugin.FlowClassifierPlugin,networking_sfc.services.sfc.plugin.SfcPlugin" fi cat >> ${local_conf_file_name} << EOF [[post-config|\$NEUTRON_CONF]] [DEFAULT] service_plugins = ${SERVICE_PLUGINS} log_dir = /opt/stack/logs [[post-config|/etc/neutron/plugins/ml2/ml2_conf.ini]] [agent] minimize_polling=True [ml2] # Needed for VLAN provider tests - because our provider networks are always encapsulated in VXLAN (br-physnet1) # MTU(1400) + VXLAN(50) + VLAN(4) = 1454 < MTU eth0/br-physnet1(1458) physical_network_mtus = ${PUBLIC_PHYSICAL_NETWORK}:1400 path_mtu = 1458 EOF if [ "${ENABLE_NETWORKING_L2GW}" == "yes" ]; then cat >> ${local_conf_file_name} << EOF [ml2_odl] enable_dhcp_service = True EOF fi cat >> ${local_conf_file_name} << EOF [ml2_odl] # Trigger n-odl full sync every 30 secs. maintenance_interval = 30 [[post-config|/etc/neutron/dhcp_agent.ini]] [DEFAULT] force_metadata = True enable_isolated_metadata = True log_dir = /opt/stack/logs [[post-config|/etc/nova/nova.conf]] [scheduler] discover_hosts_in_cells_interval = 30 [DEFAULT] force_config_drive = False force_raw_images = False log_dir = /opt/stack/logs EOF if [ "$(is_openstack_feature_enabled n-cpu)" == "1" ]; then cat >> ${local_conf_file_name} << EOF use_neutron = True force_raw_images = False log_dir = /opt/stack/logs [libvirt] live_migration_uri = qemu+tcp://%s/system virt_type = qemu EOF fi if [ "$(is_openstack_feature_enabled n-cpu)" == "1" ]; then echo "Combo local.conf created:" else echo "Control local.conf created:" fi cat ${local_conf_file_name} } # create_control_node_local_conf() function create_compute_node_local_conf() { HOSTIP=$1 SERVICEHOST=$2 MGRIP=$3 ODL_OVS_MANAGERS="$4" local_conf_file_name=${WORKSPACE}/local.conf_compute_${HOSTIP} cat > ${local_conf_file_name} << EOF [[local|localrc]] LOGFILE=stack.sh.log LOG_COLOR=False USE_SYSTEMD=True RECLONE=${RECLONE} # Increase the wait used by stack to poll for the nova service on the control node NOVA_READY_TIMEOUT=1800 disable_all_services EOF add_os_services "${CORE_OS_COMPUTE_SERVICES}" "${ENABLE_OS_COMPUTE_SERVICES}" "${DISABLE_OS_SERVICES}" "${local_conf_file_name}" cat >> ${local_conf_file_name} << EOF HOST_IP=${HOSTIP} SERVICE_HOST=${SERVICEHOST} Q_ML2_TENANT_NETWORK_TYPE=${TENANT_NETWORK_TYPE} ODL_MODE=manual ODL_MGR_IP=${MGRIP} ODL_PORT=${ODL_PORT} ODL_PORT_BINDING_CONTROLLER=${ODL_ML2_PORT_BINDING} ODL_OVS_MANAGERS=${ODL_OVS_MANAGERS} Q_HOST=\$SERVICE_HOST MYSQL_HOST=\$SERVICE_HOST RABBIT_HOST=\$SERVICE_HOST GLANCE_HOSTPORT=\$SERVICE_HOST:9292 KEYSTONE_AUTH_HOST=\$SERVICE_HOST KEYSTONE_SERVICE_HOST=\$SERVICE_HOST ADMIN_PASSWORD=${ADMIN_PASSWORD} DATABASE_PASSWORD=${ADMIN_PASSWORD} RABBIT_PASSWORD=${ADMIN_PASSWORD} SERVICE_TOKEN=${ADMIN_PASSWORD} SERVICE_PASSWORD=${ADMIN_PASSWORD} PUBLIC_BRIDGE=${PUBLIC_BRIDGE} PUBLIC_PHYSICAL_NETWORK=${PUBLIC_PHYSICAL_NETWORK} ODL_PROVIDER_MAPPINGS=${ODL_PROVIDER_MAPPINGS} EOF if [[ "${ENABLE_OS_PLUGINS}" =~ networking-odl ]]; then cat >> ${local_conf_file_name} << EOF enable_plugin networking-odl ${ODL_ML2_DRIVER_REPO} ${ODL_ML2_BRANCH} EOF fi cat >> ${local_conf_file_name} << EOF [[post-config|/etc/nova/nova.conf]] [api] auth_strategy = keystone [DEFAULT] use_neutron = True force_raw_images = False log_dir = /opt/stack/logs [libvirt] live_migration_uri = qemu+tcp://%s/system virt_type = qemu EOF echo "Compute local.conf created:" cat ${local_conf_file_name} } # create_compute_node_local_conf() function configure_haproxy_for_neutron_requests() { MGRIP=$1 # shellcheck disable=SC2206 ODL_IPS=(${2//,/ }) cat > ${WORKSPACE}/install_ha_proxy.sh<< EOF sudo systemctl stop firewalld sudo yum -y install policycoreutils-python haproxy EOF cat > ${WORKSPACE}/haproxy.cfg << EOF global daemon group haproxy log /dev/log local0 maxconn 20480 pidfile /tmp/haproxy.pid user haproxy defaults log global maxconn 4096 mode tcp retries 3 timeout http-request 10s timeout queue 1m timeout connect 10s timeout client 1m timeout server 1m timeout check 10s listen opendaylight bind ${MGRIP}:8080 balance source listen opendaylight_rest bind ${MGRIP}:8181 balance source listen opendaylight_websocket bind ${MGRIP}:8185 balance source EOF odlindex=1 for odlip in ${ODL_IPS[*]}; do sed -i "/listen opendaylight$/a server controller-${odlindex} ${odlip}:8080 check fall 5 inter 2000 rise 2" ${WORKSPACE}/haproxy.cfg sed -i "/listen opendaylight_rest$/a server controller-rest-${odlindex} ${odlip}:8181 check fall 5 inter 2000 rise 2" ${WORKSPACE}/haproxy.cfg sed -i "/listen opendaylight_websocket$/a server controller-websocket-${odlindex} ${odlip}:8185 check fall 5 inter 2000 rise 2" ${WORKSPACE}/haproxy.cfg odlindex=$((odlindex+1)) done echo "Dump haproxy.cfg" cat ${WORKSPACE}/haproxy.cfg cat > ${WORKSPACE}/deploy_ha_proxy.sh<< EOF sudo chown haproxy:haproxy /tmp/haproxy.cfg sudo sed -i 's/\\/etc\\/haproxy\\/haproxy.cfg/\\/tmp\\/haproxy.cfg/g' /usr/lib/systemd/system/haproxy.service sudo /usr/sbin/semanage permissive -a haproxy_t sudo systemctl restart haproxy sleep 3 sudo netstat -tunpl sudo systemctl status haproxy true EOF scp ${WORKSPACE}/install_ha_proxy.sh ${MGRIP}:/tmp ${SSH} ${MGRIP} "sudo bash /tmp/install_ha_proxy.sh" scp ${WORKSPACE}/haproxy.cfg ${MGRIP}:/tmp scp ${WORKSPACE}/deploy_ha_proxy.sh ${MGRIP}:/tmp ${SSH} ${MGRIP} "sudo bash /tmp/deploy_ha_proxy.sh" } # configure_haproxy_for_neutron_requests() # Following three functions are debugging helpers when debugging devstack changes. # Keeping them for now so we can simply call them when needed. ctrlhn="" comp1hn="" comp2hn="" function get_hostnames () { set +e local ctrlip=${OPENSTACK_CONTROL_NODE_1_IP} local comp1ip=${OPENSTACK_COMPUTE_NODE_1_IP} local comp2ip=${OPENSTACK_COMPUTE_NODE_2_IP} ctrlhn=$(${SSH} ${ctrlip} "hostname") comp1hn=$(${SSH} ${comp1ip} "hostname") comp2hn=$(${SSH} ${comp2ip} "hostname") echo "hostnames: ${ctrlhn}, ${comp1hn}, ${comp2hn}" set -e } function check_firewall() { set +e echo $- local ctrlip=${OPENSTACK_CONTROL_NODE_1_IP} local comp1ip=${OPENSTACK_COMPUTE_NODE_1_IP} local comp2ip=${OPENSTACK_COMPUTE_NODE_2_IP} echo "check_firewall on control" ${SSH} ${ctrlip} " sudo systemctl status firewalld sudo systemctl -l status iptables sudo iptables --line-numbers -nvL " || true echo "check_firewall on compute 1" ${SSH} ${comp1ip} " sudo systemctl status firewalld sudo systemctl -l status iptables sudo iptables --line-numbers -nvL " || true echo "check_firewall on compute 2" ${SSH} ${comp2ip} " sudo systemctl status firewalld sudo systemctl -l status iptables sudo iptables --line-numbers -nvL " || true } function get_service () { set +e local iter=$1 #local idx=$2 local ctrlip=${OPENSTACK_CONTROL_NODE_1_IP} local comp1ip=${OPENSTACK_COMPUTE_NODE_1_IP} #if [ ${idx} -eq 1 ]; then if [ ${iter} -eq 1 ] || [ ${iter} -gt 16 ]; then curl http://${ctrlip}:5000 curl http://${ctrlip}:35357 curl http://${ctrlip}/identity ${SSH} ${ctrlip} " source /opt/stack/devstack/openrc admin admin; env openstack configuration show --unmask; openstack service list openstack --os-cloud devstack-admin --os-region RegionOne compute service list openstack hypervisor list; " || true check_firewall fi #fi set -e } # Check if rabbitmq is ready by looking for a pid in it's status. # The function returns the status of the grep command which callers can check. function is_rabbitmq_ready() { local -r ip=${1} local grepfor="nova_cell1" rm -f rabbit.txt ${SSH} ${ip} "sudo rabbitmqctl list_vhosts" > rabbit.txt grep ${grepfor} rabbit.txt } # retry the given command ($3) until success for a number of iterations ($1) # sleeping ($2) between tries. function retry() { local -r -i max_tries=${1} local -r -i sleep_time=${2} local -r cmd=${3} local -i retries=1 local -i rc=1 while true; do echo "retry ${cmd}: attempt: ${retries}" ${cmd} rc=$? if ((${rc} == 0)); then break; else if ((${retries} == ${max_tries})); then break else ((retries++)) sleep ${sleep_time} fi fi done return ${rc} } function install_ovs() { local -r node=${1} local -r rpm_path=${2} if [ "${OVS_INSTALL:0:1}" = "v" ]; then # An OVS version was given, so we build it ourselves from OVS git repo. # Only on the first node though, consecutive nodes will use RPMs # built for the first one. [ ! -d "${rpm_path}" ] && mkdir -p "${rpm_path}" && build_ovs ${node} ${OVS_INSTALL} "${rpm_path}" # Install OVS from path install_ovs_from_path ${node} "${rpm_path}" elif [ "${OVS_INSTALL:0:4}" = "http" ]; then # Otherwise, install from rpm repo directly. install_ovs_from_repo ${node} ${OVS_INSTALL} else echo "Expected either an OVS version git tag or a repo http url" exit 1 fi } ODL_PROVIDER_MAPPINGS="\${PUBLIC_PHYSICAL_NETWORK}:${PUBLIC_BRIDGE}" RECLONE=False ODL_PORT=8181 # Always compare the lists below against the devstack upstream ENABLED_SERVICES in # https://github.com/openstack-dev/devstack/blob/master/stackrc#L52 # ODL CSIT does not use vnc, cinder, q-agt, q-l3 or horizon so they are not included below. # collect performance stats CORE_OS_CONTROL_SERVICES="dstat" # Glance CORE_OS_CONTROL_SERVICES+=",g-api,g-reg" # Keystone CORE_OS_CONTROL_SERVICES+=",key" # Nova - services to support libvirt CORE_OS_CONTROL_SERVICES+=",n-api,n-api-meta,n-cauth,n-cond,n-crt,n-obj,n-sch" # ODL - services to connect to ODL CORE_OS_CONTROL_SERVICES+=",odl-compute,odl-neutron" # Additional services CORE_OS_CONTROL_SERVICES+=",mysql,rabbit" # collect performance stats CORE_OS_COMPUTE_SERVICES="dstat" # computes only need nova and odl CORE_OS_COMPUTE_SERVICES+=",n-cpu,odl-compute" cat > ${WORKSPACE}/disable_firewall.sh << EOF sudo systemctl stop firewalld # Open these ports to match the tutorial vms # http/https (80/443), samba (445), netbios (137,138,139) sudo iptables -I INPUT -p tcp -m multiport --dports 80,443,139,445 -j ACCEPT sudo iptables -I INPUT -p udp -m multiport --dports 137,138 -j ACCEPT # OpenStack services as well as vxlan tunnel ports 4789 and 9876 # identity public/admin (5000/35357), ampq (5672), vnc (6080), nova (8774), glance (9292), neutron (9696) sudo sudo iptables -I INPUT -p tcp -m multiport --dports 5000,5672,6080,8774,9292,9696,35357 -j ACCEPT sudo sudo iptables -I INPUT -p udp -m multiport --dports 4789,9876 -j ACCEPT sudo iptables-save > /etc/sysconfig/iptables sudo systemctl restart iptables sudo iptables --line-numbers -nvL true EOF #For SFC Tests a larger partition is required for creating instances with Ubuntu if [[ "${ENABLE_OS_PLUGINS}" =~ networking-sfc ]]; then TMPFS_SIZE=12G fi cat > ${WORKSPACE}/get_devstack.sh << EOF sudo systemctl stop firewalld sudo yum install bridge-utils python-pip -y #sudo systemctl stop NetworkManager #Disable NetworkManager and kill dhclient and dnsmasq sudo systemctl stop NetworkManager sudo killall dhclient sudo killall dnsmasq #Workaround for mysql failure echo "127.0.0.1 localhost \${HOSTNAME}" >> /tmp/hosts echo "::1 localhost \${HOSTNAME}" >> /tmp/hosts sudo mv /tmp/hosts /etc/hosts sudo mkdir /opt/stack echo "Create RAM disk for /opt/stack" sudo mount -t tmpfs -o size=${TMPFS_SIZE} tmpfs /opt/stack sudo chmod 777 /opt/stack cd /opt/stack echo "git clone https://git.openstack.org/openstack-dev/devstack --branch ${OPENSTACK_BRANCH}" git clone https://git.openstack.org/openstack-dev/devstack --branch ${OPENSTACK_BRANCH} cd devstack if [ -n "${DEVSTACK_HASH}" ]; then echo "git checkout ${DEVSTACK_HASH}" git checkout ${DEVSTACK_HASH} fi git --no-pager log --pretty=format:'%h %<(13)%ar%<(13)%cr %<(20,trunc)%an%d %s%b' -n20 echo echo "workaround: do not upgrade openvswitch" sudo yum install -y yum-plugin-versionlock sudo yum versionlock add openvswitch #Install qemu-img command in Control Node for Pike echo "Install qemu-img application" sudo yum install -y qemu-img EOF cat > "${WORKSPACE}/setup_host_cell_mapping.sh" << EOF sudo nova-manage cell_v2 map_cell0 sudo nova-manage cell_v2 simple_cell_setup sudo nova-manage db sync sudo nova-manage cell_v2 discover_hosts EOF cat > "${WORKSPACE}/workaround_networking_sfc.sh" << EOF cd /opt/stack git clone https://git.openstack.org/openstack/networking-sfc cd networking-sfc git checkout ${OPENSTACK_BRANCH} git checkout master -- devstack/plugin.sh EOF NUM_OPENSTACK_SITES=${NUM_OPENSTACK_SITES:-1} compute_index=1 os_node_list=() if [ "${ENABLE_HAPROXY_FOR_NEUTRON}" == "yes" ]; then echo "Configure HAProxy" ODL_HAPROXYIP_PARAM=OPENSTACK_HAPROXY_1_IP ODL_IP_PARAM1=ODL_SYSTEM_1_IP ODL_IP_PARAM2=ODL_SYSTEM_2_IP ODL_IP_PARAM3=ODL_SYSTEM_3_IP ODLMGRIP=${!ODL_HAPROXYIP_PARAM} # ODL Northbound uses HAProxy VIP ODL_OVS_MGRS="${!ODL_IP_PARAM1},${!ODL_IP_PARAM2},${!ODL_IP_PARAM3}" # OVSDB connects to all ODL IPs configure_haproxy_for_neutron_requests ${!ODL_HAPROXYIP_PARAM} "${ODL_OVS_MGRS}" else ODL_IP_PARAM=ODL_SYSTEM_1_IP ODLMGRIP=${!ODL_IP_PARAM} # OVSDB connects to ODL IP ODL_OVS_MGRS="${!ODL_IP_PARAM}" # ODL Northbound uses ODL IP fi os_ip_list=() for i in `seq 1 ${NUM_OPENSTACK_CONTROL_NODES}`; do cip=OPENSTACK_CONTROL_NODE_${i}_IP ip=${!cip} os_ip_list+=("${ip}") done for i in `seq 1 ${NUM_OPENSTACK_COMPUTE_NODES}`; do cip=OPENSTACK_COMPUTE_NODE_${i}_IP ip=${!cip} os_ip_list+=("${ip}") done for i in "${!os_ip_list[@]}"; do ip=${os_ip_list[i]} tcpdump_start "${i}" "${ip}" "port 6653" done # Begin stacking the nodes, starting with the controller(s) and then the compute(s) for i in `seq 1 ${NUM_OPENSTACK_CONTROL_NODES}`; do CONTROLIP=OPENSTACK_CONTROL_NODE_${i}_IP echo "Configure the stack of the control node ${i} of ${NUM_OPENSTACK_CONTROL_NODES}: ${!CONTROLIP}" scp ${WORKSPACE}/disable_firewall.sh ${!CONTROLIP}:/tmp ${SSH} ${!CONTROLIP} "sudo bash /tmp/disable_firewall.sh" create_etc_hosts ${!CONTROLIP} scp ${WORKSPACE}/hosts_file ${!CONTROLIP}:/tmp/hosts scp ${WORKSPACE}/get_devstack.sh ${!CONTROLIP}:/tmp # devstack Master is yet to migrate fully to lib/neutron, there are some ugly hacks that is # affecting the stacking. # Workaround For Queens, Make the physical Network as physnet1 in lib/neutron # In Queens the neutron new libs are used and do not have the following options from Pike and earlier: # Q_ML2_PLUGIN_FLAT_TYPE_OPTIONS could be used for the flat_networks # and Q_ML2_PLUGIN_VLAN_TYPE_OPTIONS could be used for the ml2_type_vlan ${SSH} ${!CONTROLIP} "bash /tmp/get_devstack.sh > /tmp/get_devstack.sh.txt 2>&1" if [ "${ODL_ML2_BRANCH}" == "stable/queens" ]; then ssh ${!CONTROLIP} "sed -i 's/flat_networks public/flat_networks public,physnet1/' /opt/stack/devstack/lib/neutron" ssh ${!CONTROLIP} "sed -i '186i iniset \$NEUTRON_CORE_PLUGIN_CONF ml2_type_vlan network_vlan_ranges public:1:4094,physnet1:1:4094' /opt/stack/devstack/lib/neutron" #Workaround for networking-sfc to configure the paramaters in neutron.conf if the # services used are neutron-api, neutron-dhcp etc instead of q-agt. # Can be removed if the patch https://review.openstack.org/#/c/596287/ gets merged if [[ "${ENABLE_OS_PLUGINS}" =~ networking-sfc ]]; then scp ${WORKSPACE}/workaround_networking_sfc.sh ${!CONTROLIP}:/tmp/ ssh ${!CONTROLIP} "bash -x /tmp/workaround_networking_sfc.sh" fi fi create_control_node_local_conf ${!CONTROLIP} ${ODLMGRIP} "${ODL_OVS_MGRS}" scp ${WORKSPACE}/local.conf_control_${!CONTROLIP} ${!CONTROLIP}:/opt/stack/devstack/local.conf echo "Install rdo release to avoid incompatible Package versions" install_rdo_release ${!CONTROLIP} setup_live_migration_control ${!CONTROLIP} if [ "$(is_openstack_feature_enabled n-cpu)" == "1" ]; then setup_live_migration_compute ${!CONTROLIP} ${!CONTROLIP} fi [ -n "${OVS_INSTALL}" ] && install_ovs ${!CONTROLIP} /tmp/ovs_rpms echo "Stack the control node ${i} of ${NUM_OPENSTACK_CONTROL_NODES}: ${CONTROLIP}" ssh ${!CONTROLIP} "cd /opt/stack/devstack; nohup ./stack.sh > /opt/stack/devstack/nohup.out 2>&1 &" ssh ${!CONTROLIP} "ps -ef | grep stack.sh" ssh ${!CONTROLIP} "ls -lrt /opt/stack/devstack/nohup.out" os_node_list+=("${!CONTROLIP}") done # This is a backup to the CELLSV2_SETUP=singleconductor workaround. Keeping it here as an easy lookup # if needed. # Let the control node get started to avoid a race condition where the computes start and try to access # the nova_cell1 on the control node before it is created. If that happens, the nova-compute service on the # compute exits and does not attempt to restart. # 180s is chosen because in test runs the control node usually finished in 17-20 minutes and the computes finished # in 17 minutes, so take the max difference of 3 minutes and the jobs should still finish around the same time. # one of the following errors is seen in the compute n-cpu.log: # Unhandled error: NotAllowed: Connection.open: (530) NOT_ALLOWED - access to vhost 'nova_cell1' refused for user 'stackrabbit' # AccessRefused: (0, 0): (403) ACCESS_REFUSED - Login was refused using authentication mechanism AMQPLAIN. For details see the broker logfile. # Compare that timestamp to this log in the control stack.log: sudo rabbitmqctl set_permissions -p nova_cell1 stackrabbit # If the n-cpu.log is earlier than the control stack.log timestamp then the failure condition is likely hit. if [ ${NUM_OPENSTACK_COMPUTE_NODES} -gt 0 ]; then WAIT_FOR_RABBITMQ_MINUTES=60 echo "Wait a maximum of ${WAIT_FOR_RABBITMQ_MINUTES}m until rabbitmq is ready and nova_cell1 created to allow the controller to create nova_cell1 before the computes need it" set +e retry ${WAIT_FOR_RABBITMQ_MINUTES} 60 "is_rabbitmq_ready ${OPENSTACK_CONTROL_NODE_1_IP}" rc=$? set -e if ((${rc} == 0)); then echo "rabbitmq is ready, starting ${NUM_OPENSTACK_COMPUTE_NODES} compute(s)" else echo "rabbitmq was not ready in ${WAIT_FOR_RABBITMQ_MINUTES}m" exit 1 fi fi for i in `seq 1 ${NUM_OPENSTACK_COMPUTE_NODES}`; do COMPUTEIP=OPENSTACK_COMPUTE_NODE_${i}_IP CONTROLIP=OPENSTACK_CONTROL_NODE_1_IP echo "Configure the stack of the compute node ${i} of ${NUM_OPENSTACK_COMPUTE_NODES}: ${!COMPUTEIP}" scp ${WORKSPACE}/disable_firewall.sh "${!COMPUTEIP}:/tmp" ${SSH} "${!COMPUTEIP}" "sudo bash /tmp/disable_firewall.sh" create_etc_hosts ${!COMPUTEIP} ${!CONTROLIP} scp ${WORKSPACE}/hosts_file ${!COMPUTEIP}:/tmp/hosts scp ${WORKSPACE}/get_devstack.sh ${!COMPUTEIP}:/tmp ${SSH} ${!COMPUTEIP} "bash /tmp/get_devstack.sh > /tmp/get_devstack.sh.txt 2>&1" create_compute_node_local_conf ${!COMPUTEIP} ${!CONTROLIP} ${ODLMGRIP} "${ODL_OVS_MGRS}" scp ${WORKSPACE}/local.conf_compute_${!COMPUTEIP} ${!COMPUTEIP}:/opt/stack/devstack/local.conf echo "Install rdo release to avoid incompatible Package versions" install_rdo_release ${!COMPUTEIP} setup_live_migration_compute ${!COMPUTEIP} ${!CONTROLIP} [ -n "${OVS_INSTALL}" ] && install_ovs ${!COMPUTEIP} /tmp/ovs_rpms echo "Stack the compute node ${i} of ${NUM_OPENSTACK_COMPUTE_NODES}: ${!COMPUTEIP}" ssh ${!COMPUTEIP} "cd /opt/stack/devstack; nohup ./stack.sh > /opt/stack/devstack/nohup.out 2>&1 &" ssh ${!COMPUTEIP} "ps -ef | grep stack.sh" os_node_list+=("${!COMPUTEIP}") done echo "nodelist: ${os_node_list[*]}" # This script runs on the openstack nodes. It greps for a string that devstack writes when stacking is complete. # The script then writes a status depending on the grep output that is later scraped by the robot vm to control # the status polling. cat > ${WORKSPACE}/check_stacking.sh << EOF > /tmp/stack_progress ps -ef | grep "stack.sh" | grep -v grep ret=\$? if [ \${ret} -eq 1 ]; then grep "This is your host IP address:" /opt/stack/devstack/nohup.out if [ \$? -eq 0 ]; then echo "Stacking Complete" > /tmp/stack_progress else echo "Stacking Failed" > /tmp/stack_progress fi elif [ \${ret} -eq 0 ]; then echo "Still Stacking" > /tmp/stack_progress fi EOF # devstack debugging # get_hostnames # Check if the stacking is finished. Poll all nodes every 60s for one hour. iteration=0 in_progress=1 while [ ${in_progress} -eq 1 ]; do iteration=$(($iteration + 1)) for index in "${!os_node_list[@]}"; do echo "node $index ${os_node_list[index]}: checking stacking status attempt ${iteration} of 60" scp ${WORKSPACE}/check_stacking.sh ${os_node_list[index]}:/tmp ${SSH} ${os_node_list[index]} "bash /tmp/check_stacking.sh" scp ${os_node_list[index]}:/tmp/stack_progress . cat stack_progress stacking_status=`cat stack_progress` # devstack debugging # get_service "${iteration}" "${index}" if [ "$stacking_status" == "Still Stacking" ]; then continue elif [ "$stacking_status" == "Stacking Failed" ]; then echo "node $index ${os_node_list[index]}: stacking has failed" exit 1 elif [ "$stacking_status" == "Stacking Complete" ]; then echo "node $index ${os_node_list[index]}: stacking complete" unset 'os_node_list[index]' if [ ${#os_node_list[@]} -eq 0 ]; then in_progress=0 fi fi done echo "sleep for a minute before the next check" sleep 60 if [ ${iteration} -eq 60 ]; then echo "stacking has failed - took longer than 60m" exit 1 fi done # Further configuration now that stacking is complete. echo "Configure the Control Node" CONTROLIP=OPENSTACK_CONTROL_NODE_1_IP # Gather Compute IPs for the site for i in `seq 1 ${NUM_OPENSTACK_COMPUTE_NODES}`; do IP_VAR=OPENSTACK_COMPUTE_NODE_${i}_IP COMPUTE_IPS[$((i-1))]=${!IP_VAR} done echo "sleep for 60s and print hypervisor-list" sleep 60 ${SSH} ${!CONTROLIP} "cd /opt/stack/devstack; source openrc admin admin; nova hypervisor-list" # in the case that we are doing openstack (control + compute) all in one node, then the number of hypervisors # will be the same as the number of openstack systems. However, if we are doing multinode openstack then the # assumption is we have a single control node and the rest are compute nodes, so the number of expected hypervisors # is one less than the total number of openstack systems if [ ${NUM_OPENSTACK_SYSTEM} -eq 1 ]; then expected_num_hypervisors=1 else expected_num_hypervisors=${NUM_OPENSTACK_COMPUTE_NODES} if [ "$(is_openstack_feature_enabled n-cpu)" == "1" ]; then expected_num_hypervisors=$((expected_num_hypervisors + 1)) fi fi num_hypervisors=$(${SSH} ${!CONTROLIP} "cd /opt/stack/devstack; source openrc admin admin; openstack hypervisor list -f value | wc -l" | tail -1 | tr -d "\r") if ! [ "${num_hypervisors}" ] || ! [ ${num_hypervisors} -eq ${expected_num_hypervisors} ]; then echo "Error: Only $num_hypervisors hypervisors detected, expected $expected_num_hypervisors" exit 1 fi # External Network echo "prepare external networks by adding vxlan tunnels between all nodes on a separate bridge..." # FIXME Should there be a unique gateway IP and devstack index for each site? devstack_index=1 for ip in ${!CONTROLIP} ${COMPUTE_IPS[*]}; do # FIXME - Workaround, ODL (new netvirt) currently adds PUBLIC_BRIDGE as a port in br-int since it doesn't see such a bridge existing when we stack ${SSH} $ip "sudo ovs-vsctl --if-exists del-port br-int $PUBLIC_BRIDGE" ${SSH} $ip "sudo ovs-vsctl --may-exist add-br $PUBLIC_BRIDGE -- set bridge $PUBLIC_BRIDGE other-config:disable-in-band=true other_config:hwaddr=f6:00:00:ff:01:0$((devstack_index++))" done # ipsec support if [ "${IPSEC_VXLAN_TUNNELS_ENABLED}" == "yes" ]; then # shellcheck disable=SC2206 ALL_NODES=(${!CONTROLIP} ${COMPUTE_IPS[*]}) for ((inx_ip1=0; inx_ip1<$((${#ALL_NODES[@]} - 1)); inx_ip1++)); do for ((inx_ip2=$((inx_ip1 + 1)); inx_ip2<${#ALL_NODES[@]}; inx_ip2++)); do KEY1=0x$(dd if=/dev/urandom count=32 bs=1 2> /dev/null| xxd -p -c 64) KEY2=0x$(dd if=/dev/urandom count=32 bs=1 2> /dev/null| xxd -p -c 64) ID=0x$(dd if=/dev/urandom count=4 bs=1 2> /dev/null| xxd -p -c 8) ip1=${ALL_NODES[$inx_ip1]} ip2=${ALL_NODES[$inx_ip2]} ${SSH} $ip1 "sudo ip xfrm state add src $ip1 dst $ip2 proto esp spi $ID reqid $ID mode transport auth sha256 $KEY1 enc aes $KEY2" ${SSH} $ip1 "sudo ip xfrm state add src $ip2 dst $ip1 proto esp spi $ID reqid $ID mode transport auth sha256 $KEY1 enc aes $KEY2" ${SSH} $ip1 "sudo ip xfrm policy add src $ip1 dst $ip2 proto udp dir out tmpl src $ip1 dst $ip2 proto esp reqid $ID mode transport" ${SSH} $ip1 "sudo ip xfrm policy add src $ip2 dst $ip1 proto udp dir in tmpl src $ip2 dst $ip1 proto esp reqid $ID mode transport" ${SSH} $ip2 "sudo ip xfrm state add src $ip2 dst $ip1 proto esp spi $ID reqid $ID mode transport auth sha256 $KEY1 enc aes $KEY2" ${SSH} $ip2 "sudo ip xfrm state add src $ip1 dst $ip2 proto esp spi $ID reqid $ID mode transport auth sha256 $KEY1 enc aes $KEY2" ${SSH} $ip2 "sudo ip xfrm policy add src $ip2 dst $ip1 proto udp dir out tmpl src $ip2 dst $ip1 proto esp reqid $ID mode transport" ${SSH} $ip2 "sudo ip xfrm policy add src $ip1 dst $ip2 proto udp dir in tmpl src $ip1 dst $ip2 proto esp reqid $ID mode transport" done done for ip in ${!CONTROLIP} ${COMPUTE_IPS[*]}; do echo "ip xfrm configuration for node $ip:" ${SSH} $ip "sudo ip xfrm policy list" ${SSH} $ip "sudo ip xfrm state list" done fi # Control Node - PUBLIC_BRIDGE will act as the external router # Parameter values below are used in integration/test - changing them requires updates in intergration/test as well EXTNET_GATEWAY_IP="10.10.10.250" EXTNET_INTERNET_IP="10.9.9.9" EXTNET_PNF_IP="10.10.10.253" ${SSH} ${!CONTROLIP} "sudo ifconfig ${PUBLIC_BRIDGE} up ${EXTNET_GATEWAY_IP}/24" # Control Node - external net PNF simulation ${SSH} ${!CONTROLIP} " sudo ip netns add pnf_ns; sudo ip link add pnf_veth0 type veth peer name pnf_veth1; sudo ip link set pnf_veth1 netns pnf_ns; sudo ip link set pnf_veth0 up; sudo ip netns exec pnf_ns ifconfig pnf_veth1 up ${EXTNET_PNF_IP}/24; sudo ovs-vsctl add-port ${PUBLIC_BRIDGE} pnf_veth0; " # Control Node - external net internet address simulation ${SSH} ${!CONTROLIP} " sudo ip tuntap add dev internet_tap mode tap; sudo ifconfig internet_tap up ${EXTNET_INTERNET_IP}/24; " # Computes compute_index=1 for compute_ip in ${COMPUTE_IPS[*]}; do # Tunnel from controller to compute COMPUTEPORT=compute$(( compute_index++ ))_vxlan ${SSH} ${!CONTROLIP} " sudo ovs-vsctl add-port $PUBLIC_BRIDGE $COMPUTEPORT -- set interface $COMPUTEPORT type=vxlan options:local_ip=${!CONTROLIP} options:remote_ip=$compute_ip options:dst_port=9876 options:key=flow " # Tunnel from compute to controller CONTROLPORT="control_vxlan" ${SSH} $compute_ip " sudo ovs-vsctl add-port $PUBLIC_BRIDGE $CONTROLPORT -- set interface $CONTROLPORT type=vxlan options:local_ip=$compute_ip options:remote_ip=${!CONTROLIP} options:dst_port=9876 options:key=flow " done if [ "${ENABLE_HAPROXY_FOR_NEUTRON}" == "yes" ]; then odlmgrip=OPENSTACK_HAPROXY_1_IP HA_PROXY_IP=${!odlmgrip} HA_PROXY_1_IP=${!odlmgrip} odlmgrip2=OPENSTACK_HAPROXY_2_IP HA_PROXY_2_IP=${!odlmgrip2} odlmgrip3=OPENSTACK_HAPROXY_1_IP HA_PROXY_3_IP=${!odlmgrip3} else HA_PROXY_IP=${ODL_SYSTEM_IP} HA_PROXY_1_IP=${ODL_SYSTEM_1_IP} HA_PROXY_2_IP=${ODL_SYSTEM_2_IP} HA_PROXY_3_IP=${ODL_SYSTEM_3_IP} fi echo "Locating test plan to use..." testplan_filepath="${WORKSPACE}/test/csit/testplans/${STREAMTESTPLAN}" if [ ! -f "${testplan_filepath}" ]; then testplan_filepath="${WORKSPACE}/test/csit/testplans/${TESTPLAN}" fi echo "Changing the testplan path..." cat "${testplan_filepath}" | sed "s:integration:${WORKSPACE}:" > testplan.txt cat testplan.txt # Use the testplan if specific SUITES are not defined. if [ -z "${SUITES}" ]; then SUITES=`egrep -v '(^[[:space:]]*#|^[[:space:]]*$)' testplan.txt | tr '\012' ' '` else newsuites="" workpath="${WORKSPACE}/test/csit/suites" for suite in ${SUITES}; do fullsuite="${workpath}/${suite}" if [ -z "${newsuites}" ]; then newsuites+=${fullsuite} else newsuites+=" "${fullsuite} fi done SUITES=${newsuites} fi #install all client versions required for this job testing install_openstack_clients_in_robot_vm # TODO: run openrc on control node and then scrape the vars from it # Environment Variables Needed to execute Openstack Client for NetVirt Jobs cat > /tmp/os_netvirt_client_rc << EOF export OS_USERNAME=admin export OS_PASSWORD=admin export OS_PROJECT_NAME=admin export OS_USER_DOMAIN_NAME=default export OS_PROJECT_DOMAIN_NAME=default export OS_AUTH_URL="http://${!CONTROLIP}/identity" export OS_IDENTITY_API_VERSION=3 export OS_IMAGE_API_VERSION=2 export OS_TENANT_NAME=admin unset OS_CLOUD EOF source /tmp/os_netvirt_client_rc echo "Get all versions before executing pybot" echo "openstack --version" which openstack openstack --version echo "nova --version" which nova nova --version echo "neutron --version" which neutron neutron --version stacktime=$(timer $totaltmr) printf "Stacking elapsed time: %s\n" "${stacktime}" echo "Starting Robot test suites ${SUITES} ..." # please add pybot -v arguments on a single line and alphabetized suite_num=0 for suite in ${SUITES}; do # prepend an incremental counter to the suite name so that the full robot log combining all the suites as is done # in the rebot step below will list all the suites in chronological order as rebot seems to alphabetize them let "suite_num = suite_num + 1" suite_index="$(printf %02d ${suite_num})" suite_name="$(basename ${suite} | cut -d. -f1)" log_name="${suite_index}_${suite_name}" pybot -N ${log_name} \ -c critical -e exclude -e skip_if_${DISTROSTREAM} \ --log log_${log_name}.html --report report_${log_name}.html --output output_${log_name}.xml \ --removekeywords wuks \ --removekeywords name:SetupUtils.Setup_Utils_For_Setup_And_Teardown \ --removekeywords name:SetupUtils.Setup_Test_With_Logging_And_Without_Fast_Failing \ --removekeywords name:OpenStackOperations.Add_OVS_Logging_On_All_OpenStack_Nodes \ -v BUNDLEFOLDER:${BUNDLEFOLDER} \ -v BUNDLE_URL:${ACTUAL_BUNDLE_URL} \ -v CMP_INSTANCES_SHARED_PATH:/var/instances \ -v CONTROLLERFEATURES:"${CONTROLLERFEATURES}" \ -v CONTROLLER_USER:${USER} \ -v DEVSTACK_DEPLOY_PATH:/opt/stack/devstack \ -v ENABLE_ITM_DIRECT_TUNNELS:${ENABLE_ITM_DIRECT_TUNNELS} \ -v HA_PROXY_IP:${HA_PROXY_IP} \ -v HA_PROXY_1_IP:${HA_PROXY_1_IP} \ -v HA_PROXY_2_IP:${HA_PROXY_2_IP} \ -v HA_PROXY_3_IP:${HA_PROXY_3_IP} \ -v JDKVERSION:${JDKVERSION} \ -v JENKINS_WORKSPACE:${WORKSPACE} \ -v NEXUSURL_PREFIX:${NEXUSURL_PREFIX} \ -v NUM_ODL_SYSTEM:${NUM_ODL_SYSTEM} \ -v NUM_OS_SYSTEM:${NUM_OPENSTACK_SYSTEM} \ -v NUM_TOOLS_SYSTEM:${NUM_TOOLS_SYSTEM} \ -v ODL_SNAT_MODE:${ODL_SNAT_MODE} \ -v ODL_STREAM:${DISTROSTREAM} \ -v ODL_SYSTEM_IP:${ODL_SYSTEM_IP} \ -v ODL_SYSTEM_1_IP:${ODL_SYSTEM_1_IP} \ -v ODL_SYSTEM_2_IP:${ODL_SYSTEM_2_IP} \ -v ODL_SYSTEM_3_IP:${ODL_SYSTEM_3_IP} \ -v ODL_SYSTEM_4_IP:${ODL_SYSTEM_4_IP} \ -v ODL_SYSTEM_5_IP:${ODL_SYSTEM_5_IP} \ -v ODL_SYSTEM_6_IP:${ODL_SYSTEM_6_IP} \ -v ODL_SYSTEM_7_IP:${ODL_SYSTEM_7_IP} \ -v ODL_SYSTEM_8_IP:${ODL_SYSTEM_8_IP} \ -v ODL_SYSTEM_9_IP:${ODL_SYSTEM_9_IP} \ -v OS_CONTROL_NODE_IP:${OPENSTACK_CONTROL_NODE_1_IP} \ -v OS_CONTROL_NODE_1_IP:${OPENSTACK_CONTROL_NODE_1_IP} \ -v OS_CONTROL_NODE_2_IP:${OPENSTACK_CONTROL_NODE_2_IP} \ -v OS_CONTROL_NODE_3_IP:${OPENSTACK_CONTROL_NODE_3_IP} \ -v OPENSTACK_BRANCH:${OPENSTACK_BRANCH} \ -v OS_COMPUTE_1_IP:${OPENSTACK_COMPUTE_NODE_1_IP} \ -v OS_COMPUTE_2_IP:${OPENSTACK_COMPUTE_NODE_2_IP} \ -v OS_COMPUTE_3_IP:${OPENSTACK_COMPUTE_NODE_3_IP} \ -v OS_COMPUTE_4_IP:${OPENSTACK_COMPUTE_NODE_4_IP} \ -v OS_COMPUTE_5_IP:${OPENSTACK_COMPUTE_NODE_5_IP} \ -v OS_COMPUTE_6_IP:${OPENSTACK_COMPUTE_NODE_6_IP} \ -v OPENSTACK_TOPO:${OPENSTACK_TOPO} \ -v OS_USER:${USER} \ -v PUBLIC_PHYSICAL_NETWORK:${PUBLIC_PHYSICAL_NETWORK} \ -v SECURITY_GROUP_MODE:${SECURITY_GROUP_MODE} \ -v TOOLS_SYSTEM_IP:${TOOLS_SYSTEM_1_IP} \ -v TOOLS_SYSTEM_1_IP:${TOOLS_SYSTEM_1_IP} \ -v TOOLS_SYSTEM_2_IP:${TOOLS_SYSTEM_2_IP} \ -v USER_HOME:${HOME} \ -v WORKSPACE:/tmp \ ${TESTOPTIONS} ${suite} || true done #rebot exit codes seem to be different rebot --output ${WORKSPACE}/output.xml --log log_full.html --report report.html -N openstack output_*.xml || true echo "Examining the files in data/log and checking file size" ssh ${ODL_SYSTEM_IP} "ls -altr /tmp/${BUNDLEFOLDER}/data/log/" ssh ${ODL_SYSTEM_IP} "du -hs /tmp/${BUNDLEFOLDER}/data/log/*" echo "Tests Executed" printf "Total elapsed time: %s, stacking time: %s\n" "$(timer $totaltmr)" "${stacktime}" true # perhaps Jenkins is testing last exit code # vim: ts=4 sw=4 sts=4 et ft=sh :