diff options
Diffstat (limited to 'extraconfig')
| -rwxr-xr-x | extraconfig/tasks/pacemaker_common_functions.sh | 49 | ||||
| -rw-r--r-- | extraconfig/tasks/post_puppet_pacemaker.j2.yaml (renamed from extraconfig/tasks/post_puppet_pacemaker.yaml) | 21 | ||||
| -rw-r--r-- | extraconfig/tasks/post_puppet_pacemaker_restart.yaml | 2 | ||||
| -rw-r--r-- | extraconfig/tasks/pre_puppet_pacemaker.yaml | 2 | ||||
| -rwxr-xr-x | extraconfig/tasks/yum_update.sh | 48 |
5 files changed, 89 insertions, 33 deletions
diff --git a/extraconfig/tasks/pacemaker_common_functions.sh b/extraconfig/tasks/pacemaker_common_functions.sh index 4480f74d..f17a073a 100755 --- a/extraconfig/tasks/pacemaker_common_functions.sh +++ b/extraconfig/tasks/pacemaker_common_functions.sh @@ -322,3 +322,52 @@ function special_case_ovs_upgrade_if_needed { } +# This code is meant to fix https://bugs.launchpad.net/tripleo/+bug/1686357 on +# existing setups via a minor update workflow and be idempotent. We need to +# run this before the yum update because we fix this up even when there are no +# packages to update on the system (in which case the script exits). +# This code must be called with set +eu (due to the ocf scripts being sourced) +function fixup_wrong_ipv6_vip { + # This XPath query identifies of all the VIPs in pacemaker with netmask /64. Those are IPv6 only resources that have the wrong netmask + # This gives the address of the resource in the CIB, one address per line. For example: + # /cib/configuration/resources/primitive[@id='ip-2001.db8.ca2.4..10']/instance_attributes[@id='ip-2001.db8.ca2.4..10-instance_attributes']\ + # /nvpair[@id='ip-2001.db8.ca2.4..10-instance_attributes-cidr_netmask'] + vip_xpath_query="//resources/primitive[@type='IPaddr2']/instance_attributes/nvpair[@name='cidr_netmask' and @value='64']" + vip_xpath_xml_addresses=$(cibadmin --query --xpath "$vip_xpath_query" -e 2>/dev/null) + # The following extracts the @id value of the resource + vip_resources_to_fix=$(echo -e "$vip_xpath_xml_addresses" | sed -n "s/.*primitive\[@id='\([^']*\)'.*/\1/p") + # Runnning this in a subshell so that sourcing files cannot possibly affect the running script + ( + OCF_PATH="/usr/lib/ocf/lib/heartbeat" + if [ -n "$vip_resources_to_fix" -a -f $OCF_PATH/ocf-shellfuncs -a -f $OCF_PATH/findif.sh ]; then + source $OCF_PATH/ocf-shellfuncs + source $OCF_PATH/findif.sh + for resource in $vip_resources_to_fix; do + echo "Updating IPv6 VIP $resource with a /128 and a correct addrlabel" + # The following will give us something like: + # <nvpair id="ip-2001.db8.ca2.4..10-instance_attributes-ip" name="ip" value="2001:db8:ca2:4::10"/> + ip_cib_nvpair=$(cibadmin --query --xpath "//resources/primitive[@type='IPaddr2' and @id='$resource']/instance_attributes/nvpair[@name='ip']") + # Let's filter out the value of the nvpair to get the ip address + ip_address=$(echo $ip_cib_nvpair | xmllint --xpath 'string(//nvpair/@value)' -) + OCF_RESKEY_cidr_netmask="64" + OCF_RESKEY_ip="$ip_address" + # Unfortunately due to https://bugzilla.redhat.com/show_bug.cgi?id=1445628 + # we need to find out the appropiate nic given the ip address. + nic=$(findif $ip_address | awk '{ print $1 }') + ret=$? + if [ -z "$nic" -o $ret -ne 0 ]; then + echo "NIC autodetection failed for VIP $ip_address, not updating VIPs" + # Only exits the subshell + exit 1 + fi + ocf_run -info pcs resource update --wait "$resource" ip="$ip_address" cidr_netmask=128 nic="$nic" lvs_ipv6_addrlabel=true lvs_ipv6_addrlabel_value=99 + ret=$? + if [ $ret -ne 0 ]; then + echo "pcs resource update for VIP $resource failed, not updating VIPs" + # Only exits the subshell + exit 1 + fi + done + fi + ) +} diff --git a/extraconfig/tasks/post_puppet_pacemaker.yaml b/extraconfig/tasks/post_puppet_pacemaker.j2.yaml index a304e55b..0db0bc12 100644 --- a/extraconfig/tasks/post_puppet_pacemaker.yaml +++ b/extraconfig/tasks/post_puppet_pacemaker.j2.yaml @@ -10,7 +10,9 @@ parameters: resources: - ControllerPostPuppetMaintenanceModeConfig: +{%- for role in roles -%} +{% if "controller" in role.tags %} + {{role.name}}PostPuppetMaintenanceModeConfig: type: OS::Heat::SoftwareConfig properties: group: script @@ -22,16 +24,19 @@ resources: pcs property set maintenance-mode=false fi - ControllerPostPuppetMaintenanceModeDeployment: + {{role.name}}PostPuppetMaintenanceModeDeployment: type: OS::Heat::SoftwareDeployments properties: - servers: {get_param: servers} - config: {get_resource: ControllerPostPuppetMaintenanceModeConfig} + servers: {get_param: [servers, {{role.name}}]} + config: {get_resource: {{role.name}}PostPuppetMaintenanceModeConfig} input_values: {get_param: input_values} - ControllerPostPuppetRestart: - type: OS::TripleO::Tasks::ControllerPostPuppetRestart - depends_on: ControllerPostPuppetMaintenanceModeDeployment + {{role.name}}PostPuppetRestart: + type: OS::TripleO::Tasks::{{role.name}}PostPuppetRestart + depends_on: {{role.name}}PostPuppetMaintenanceModeDeployment properties: - servers: {get_param: servers} + servers: {get_param: [servers, {{role.name}}]} input_values: {get_param: input_values} +{%- endif -%} +{% endfor %} + diff --git a/extraconfig/tasks/post_puppet_pacemaker_restart.yaml b/extraconfig/tasks/post_puppet_pacemaker_restart.yaml index 475a6688..07f3d00d 100644 --- a/extraconfig/tasks/post_puppet_pacemaker_restart.yaml +++ b/extraconfig/tasks/post_puppet_pacemaker_restart.yaml @@ -23,6 +23,6 @@ resources: ControllerPostPuppetRestartDeployment: type: OS::Heat::SoftwareDeployments properties: - servers: {get_param: servers} + servers: {get_param: servers} config: {get_resource: ControllerPostPuppetRestartConfig} input_values: {get_param: input_values} diff --git a/extraconfig/tasks/pre_puppet_pacemaker.yaml b/extraconfig/tasks/pre_puppet_pacemaker.yaml index aa7514f9..a4244633 100644 --- a/extraconfig/tasks/pre_puppet_pacemaker.yaml +++ b/extraconfig/tasks/pre_puppet_pacemaker.yaml @@ -20,6 +20,6 @@ resources: ControllerPrePuppetMaintenanceModeDeployment: type: OS::Heat::SoftwareDeployments properties: - servers: {get_param: servers} + servers: {get_param: servers} config: {get_resource: ControllerPrePuppetMaintenanceModeConfig} input_values: {get_param: input_values} diff --git a/extraconfig/tasks/yum_update.sh b/extraconfig/tasks/yum_update.sh index 20a5b658..83d6d8d6 100755 --- a/extraconfig/tasks/yum_update.sh +++ b/extraconfig/tasks/yum_update.sh @@ -38,6 +38,29 @@ if [[ -a "$timestamp_file" ]]; then fi touch "$timestamp_file" +pacemaker_status="" +if hiera -c /etc/puppet/hiera.yaml service_names | grep -q pacemaker; then + pacemaker_status=$(systemctl is-active pacemaker) +fi + +# (NB: when backporting this s/pacemaker_short_bootstrap_node_name/bootstrap_nodeid) +# This runs before the yum_update so we are guaranteed to run it even in the absence +# of packages to update (the check for -z "$update_identifier" guarantees that this +# is run only on overcloud stack update -i) +if [[ "$pacemaker_status" == "active" && \ + "$(hiera -c /etc/puppet/hiera.yaml pacemaker_short_bootstrap_node_name)" == "$(facter hostname)" ]] ; then \ + # OCF scripts don't cope with -eu + echo "Verifying if we need to fix up any IPv6 VIPs" + set +eu + fixup_wrong_ipv6_vip + ret=$? + set -eu + if [ $ret -ne 0 ]; then + echo "Fixing up IPv6 VIPs failed. Stopping here. (See https://bugs.launchpad.net/tripleo/+bug/1686357 for more info)" + exit 1 + fi +fi + command_arguments=${command_arguments:-} # yum check-update exits 100 if updates are available @@ -55,28 +78,6 @@ elif [[ "$check_update_exit" != "100" ]]; then exit 0 fi -pacemaker_status="" -if hiera -c /etc/puppet/hiera.yaml service_names | grep -q pacemaker; then - pacemaker_status=$(systemctl is-active pacemaker) -fi - -# Fix the redis/rabbit resource start/stop timeouts. See https://bugs.launchpad.net/tripleo/+bug/1633455 -# and https://bugs.launchpad.net/tripleo/+bug/1634851 -if [[ "$pacemaker_status" == "active" && \ - "$(hiera -c /etc/puppet/hiera.yaml bootstrap_nodeid)" = "$(facter hostname)" ]] ; then - if pcs resource show rabbitmq | grep -E "start.*timeout=100"; then - pcs resource update rabbitmq op start timeout=200s - fi - if pcs resource show rabbitmq | grep -E "stop.*timeout=90"; then - pcs resource update rabbitmq op stop timeout=200s - fi - if pcs resource show redis | grep -E "start.*timeout=120"; then - pcs resource update redis op start timeout=200s - fi - if pcs resource show redis | grep -E "stop.*timeout=120"; then - pcs resource update redis op stop timeout=200s - fi -fi # special case https://bugs.launchpad.net/tripleo/+bug/1635205 +bug/1669714 special_case_ovs_upgrade_if_needed @@ -147,6 +148,7 @@ if [[ "$pacemaker_status" == "active" ]] ; then pcs status fi -echo "Finished yum_update.sh on server $deploy_server_id at `date`" + +echo "Finished yum_update.sh on server $deploy_server_id at `date` with return code: $return_code" exit $return_code |