tcp_tests/templates/cookied-mcp-ocata-dop-sl2/oss.yaml

{% from 'cookied-mcp-ocata-dop-sl2/underlay.yaml' import HOSTNAME_CFG01 with context %}

# Install OSS: Operational Support System Tools

# Keepalived
#-----------
- description: Install keepalived
  cmd: salt --hard-crash --state-output=mixed --state-verbose=False -C 'I@keepalived:cluster:enabled:True' state.sls keepalived
  node_name: {{ HOSTNAME_CFG01 }}
  retry: {count: 1, delay: 10}
  skip_fail: false

- description: Install haproxy
  cmd: salt --hard-crash --state-output=mixed --state-verbose=False -C 'I@haproxy:proxy:enabled:True' state.sls haproxy
  node_name: {{ HOSTNAME_CFG01 }}
  retry: {count: 1, delay: 10}
  skip_fail: false

- description: Check the CICD VIP
  cmd: |
    CICD_CONTROL_ADDRESS=`salt --out=newline_values_only -C 'I@haproxy:proxy and I@jenkins:client' pillar.get _param:cluster_vip_address`;
    echo "_param:cluster_vip_address (vip): ${CICD_CONTROL_ADDRESS}";
    salt --hard-crash --state-output=mixed --state-verbose=False -C "I@keepalived:cluster:instance:*:address:${CICD_CONTROL_ADDRESS}" cmd.run "ip a | grep ${CICD_CONTROL_ADDRESS}" | grep -B1 ${CICD_CONTROL_ADDRESS}
  node_name: {{ HOSTNAME_CFG01 }}
  retry: {count: 1, delay: 5}
  skip_fail: false

# Glusterfs
#-----------

- description: Prepare glusterfs service
  cmd: salt --hard-crash --state-output=mixed --state-verbose=False -C 'I@glusterfs:server:enabled:True' state.sls glusterfs.server.service
  node_name: {{ HOSTNAME_CFG01 }}
  retry: {count: 1, delay: 5}
  skip_fail: false

- description: Setup glusterfs server
  cmd: salt --hard-crash --state-output=mixed --state-verbose=False -C 'I@glusterfs:server:enabled:True' state.sls glusterfs.server.setup -b 1
  node_name: {{ HOSTNAME_CFG01 }}
  retry: {count: 2, delay: 5}
  skip_fail: false

- description: Setup glusterfs client
  cmd: salt --hard-crash --state-output=mixed --state-verbose=False -C 'I@glusterfs:client:enabled:True' state.sls glusterfs.client
  node_name: {{ HOSTNAME_CFG01 }}
  retry: {count: 1, delay: 5}
  skip_fail: false

- description: Check the gluster status
  cmd: salt --hard-crash --state-output=mixed --state-verbose=False
    -C 'I@glusterfs:server:enabled:True' cmd.run 'gluster peer status; gluster volume status' -b 1
  node_name: {{ HOSTNAME_CFG01 }}
  retry: {count: 1, delay: 5}
  skip_fail: false

# Elasticsearch (system service)
#-------------------------------
- description: Setup Elasticsearch
  cmd: salt --hard-crash --state-output=mixed --state-verbose=False -C 'I@elasticsearch:server' state.sls elasticsearch.server -b 1
  node_name: {{ HOSTNAME_CFG01 }}
  retry: {count: 3, delay: 10}
  skip_fail: false

- description: Setup Elasticsearch
  cmd: salt --hard-crash --state-output=mixed --state-verbose=False -C 'I@elasticsearch:client' state.sls elasticsearch.client
  node_name: {{ HOSTNAME_CFG01 }}
  retry: {count: 3, delay: 10}
  skip_fail: false

# Setup Docker Swarm
#-------------------

- description: "Workaround: create /var/lib/jenkins to get Jenkins slaves working"
  cmd: salt --hard-crash --state-output=mixed --state-verbose=False -C 'I@jenkins:client' cmd.run 'mkdir -p /var/lib/jenkins'
  node_name: {{ HOSTNAME_CFG01 }}
  retry: {count: 1, delay: 10}
  skip_fail: false

- description: Prepare Docker host
  cmd: salt --hard-crash --state-output=mixed --state-verbose=False -C 'I@docker:host:enabled:True' state.sls docker.host
  node_name: {{ HOSTNAME_CFG01 }}
  retry: {count: 1, delay: 5}
  skip_fail: false

- description: Install Docker Swarm master
  cmd: salt --hard-crash --state-output=mixed --state-verbose=False -C 'I@docker:swarm:role:master' state.sls docker.swarm
  node_name: {{ HOSTNAME_CFG01 }}
  retry: {count: 1, delay: 5}
  skip_fail: false

- description: Collect grains
  cmd: salt --hard-crash --state-output=mixed --state-verbose=False -C 'I@docker:swarm:role:master' state.sls salt.minion.grains &&
    salt --hard-crash --state-output=mixed --state-verbose=False -C 'I@docker:swarm:role:master' mine.flush &&
    salt --hard-crash --state-output=mixed --state-verbose=False -C 'I@docker:swarm:role:master' mine.update &&
    salt --hard-crash --state-output=mixed --state-verbose=False -C 'I@docker:swarm' saltutil.refresh_modules &&
    sleep 10
  node_name: {{ HOSTNAME_CFG01 }}
  retry: {count: 3, delay: 15}
  skip_fail: false

- description: Install Docker Swarm on other nodes
  cmd: salt --hard-crash --state-output=mixed --state-verbose=False -C 'I@docker:swarm' state.sls docker.swarm
  node_name: {{ HOSTNAME_CFG01 }}
  retry: {count: 2, delay: 5}
  skip_fail: false

- description: Show Docker Swarm nodes
  cmd: salt --hard-crash --state-output=mixed --state-verbose=False -C 'I@docker:swarm:role:master' cmd.run 'docker node ls'
  node_name: {{ HOSTNAME_CFG01 }}
  retry: {count: 1, delay: 5}
  skip_fail: false

# Configure OSS services
#-----------------------

- description: Setup devops portal
  cmd: salt --hard-crash --state-output=mixed --state-verbose=False -C 'I@devops_portal:config:enabled' state.sls devops_portal.config
  node_name: {{ HOSTNAME_CFG01 }}
  retry: {count: 1, delay: 5}
  skip_fail: false

- description: Setup Rundeck server
  cmd: salt --hard-crash --state-output=mixed --state-verbose=False -C 'I@rundeck:server' state.sls rundeck.server
  node_name: {{ HOSTNAME_CFG01 }}
  retry: {count: 1, delay: 5}
  skip_fail: false

# Deploy Docker services
#-----------------------

# Original comment from pipeline:  XXX: for some weird unknown reason, refresh_pillar is required to execute here

- description: "Workaround from the pipeline: XXX: for some weird unknown reason, refresh_pillar is required to execute here"
  cmd: salt --hard-crash --state-output=mixed --state-verbose=False -C 'I@aptly:publisher' saltutil.refresh_pillar
  node_name: {{ HOSTNAME_CFG01 }}
  retry: {count: 1, delay: 5}
  skip_fail: false

- description: "Workaround from the pipeline: We need /etc/aptly-publisher.yaml to be present before services are deployed. [dd: there were issues when /etc/aptly-publisher.yaml becomes a directory, so this step should be considered]"
  cmd: salt --hard-crash --state-output=mixed --state-verbose=False -C 'I@aptly:publisher' state.sls aptly.publisher
  node_name: {{ HOSTNAME_CFG01 }}
  retry: {count: 2, delay: 5}
  skip_fail: false

- description: Install Docker client
  cmd: salt --hard-crash --state-output=mixed --state-verbose=False -C 'I@docker:swarm:role:master' state.sls docker.client
  node_name: {{ HOSTNAME_CFG01 }}
  retry: {count: 3, delay: 5}
  skip_fail: false

- description: "Workaround from the pipeline: sync all salt objects"
  cmd: salt --hard-crash --state-output=mixed --state-verbose=False '*' saltutil.sync_all && sleep 5
  node_name: {{ HOSTNAME_CFG01 }}
  retry: {count: 1, delay: 5}
  skip_fail: false


# Aptly
#------

#### Steps are commented due to PROD-17598

#- description: "Wait for Aptly to come up in container..."
#  cmd: timeout 300 salt --hard-crash --state-output=mixed --state-verbose=False -C 'I@aptly:server' cmd.run
#      'export CICD_CONTROL_ADDRESS=$(salt-call --out=newline_values_only pillar.get _param:cluster_vip_address);
#       while true; do curl -sf http://${CICD_CONTROL_ADDRESS}:8084/api/version  && break; sleep 2; done'
#  node_name: {{ HOSTNAME_CFG01 }}
#  retry: {count: 3, delay: 15}
#  skip_fail: false

#- description: "Setup Aptly"
#  cmd: salt --hard-crash --state-output=mixed --state-verbose=False -C 'I@aptly:server' state.sls aptly
#  node_name: {{ HOSTNAME_CFG01 }}
#  retry: {count: 3, delay: 10}
#  skip_fail: false

# OpenLDAP
#---------

- description: "Waiting for OpenLDAP to come up in container..."
  cmd: timeout 60 salt --hard-crash --state-output=mixed --state-verbose=False -C 'I@openldap:client' cmd.run
      'export CICD_CONTROL_ADDRESS=$(salt-call --out=newline_values_only pillar.get _param:cluster_vip_address);
       while true; do curl -sf ldap://${CICD_CONTROL_ADDRESS} && break; sleep 2; done'
  node_name: {{ HOSTNAME_CFG01 }}
  retry: {count: 3, delay: 10}
  skip_fail: false

- description: "Setup OpenLDAP"
  cmd: salt --hard-crash --state-output=mixed --state-verbose=False -C 'I@openldap:client' state.sls openldap &&
    sleep 20
  node_name: {{ HOSTNAME_CFG01 }}
  retry: {count: 3, delay: 10}
  skip_fail: false

# Gerrit
#-------

- description: "Waiting for Gerrit to come up in container..."
  cmd: timeout 60 salt --hard-crash --state-output=mixed --state-verbose=False -C 'I@gerrit:client' cmd.run
      'export CICD_CONTROL_ADDRESS=$(salt-call --out=newline_values_only pillar.get _param:cluster_vip_address);
       while true; do curl -sf http://${CICD_CONTROL_ADDRESS}:8080/config/server/version && break; sleep 2; done'
  node_name: {{ HOSTNAME_CFG01 }}
  retry: {count: 3, delay: 10}
  skip_fail: false

- description: "Setup Gerrit"
  cmd: salt --hard-crash --state-output=mixed --state-verbose=False -C 'I@gerrit:client' state.sls gerrit
  node_name: {{ HOSTNAME_CFG01 }}
  retry: {count: 3, delay: 10}
  skip_fail: false

# Jenkins
#--------

- description: "Waiting for Jenkins to come up in container..."
  cmd: timeout 60 salt --hard-crash --state-output=mixed --state-verbose=False -C 'I@jenkins:client' cmd.run
      'export CICD_CONTROL_ADDRESS=$(salt-call --out=newline_values_only pillar.get _param:cluster_vip_address);
       export JENKINS_CLIENT_USER=$(salt-call --out=newline_values_only pillar.get _param:jenkins_client_user);
       export JENKINS_CLIENT_PASSWORD=$(salt-call --out=newline_values_only pillar.get _param:jenkins_client_password);
       while true; do
         curl -f -u ${JENKINS_CLIENT_USER}:${JENKINS_CLIENT_PASSWORD} http://${CICD_CONTROL_ADDRESS}:8081/api/json?pretty=true && break;
         sleep 2;
       done'
  node_name: {{ HOSTNAME_CFG01 }}
  retry: {count: 3, delay: 10}
  skip_fail: false

- description: "Setup Jenkins"
  cmd: salt --hard-crash --state-output=mixed --state-verbose=False -C 'I@jenkins:client' state.sls jenkins
  node_name: {{ HOSTNAME_CFG01 }}
  retry: {count: 10, delay: 30}
  skip_fail: false

# Postgres && Pushkin
#--------------------

- description: "Waiting for postgresql database to come up in container..."
#  cmd: timeout 300 salt --hard-crash --state-output=mixed --state-verbose=False -C 'I@postgresql:client' cmd.run
#      'while true; do if docker service logs postgresql_db | grep -q "ready to accept"; then break; else sleep 5; fi; done'
  cmd: timeout 300 salt --hard-crash --state-output=mixed --state-verbose=False -C 'I@postgresql:client' cmd.run
      'while true; do if docker service logs postgresql_postgresql-db | grep -q "ready to accept"; then break; else sleep 5; fi; done'
  node_name: {{ HOSTNAME_CFG01 }}
  retry: {count: 3, delay: 10}
  skip_fail: false

- description: ("Create PostgreSQL databases, it fails at first run because of known deadlock:\n"
                "1. State postgresql.client cannot insert values into 'pushkin' database because it is created empty,\n"
                "2. Container with Pushkin cannot start and fill the database scheme until state postgresql.client created users.")
  cmd: salt --hard-crash --state-output=mixed --state-verbose=False -C 'I@postgresql:client' state.sls postgresql.client -b 1 &&
    timeout 300 salt --hard-crash --state-output=mixed --state-verbose=False -C 'I@postgresql:client' cmd.run
    'export CICD_CONTROL_ADDRESS=$(salt-call --out=newline_values_only pillar.get _param:cluster_vip_address);
     while true; do curl -sf http://${CICD_CONTROL_ADDRESS}:8887/apps && break; sleep 2; done'
  node_name: {{ HOSTNAME_CFG01 }}
  retry: {count: 3, delay: 100}
  skip_fail: false

# Rundeck
#--------

- description: Waiting for Rundeck to come up in container...
  cmd: timeout 30 salt --hard-crash --state-output=mixed --state-verbose=False -C 'I@rundeck:client' cmd.run
      'export CICD_CONTROL_ADDRESS=$(salt-call --out=newline_values_only pillar.get _param:cluster_vip_address);
       while true; do curl -sf http://${CICD_CONTROL_ADDRESS}:4440 && break; sleep 2; done'
  node_name: {{ HOSTNAME_CFG01 }}
  retry: {count: 3, delay: 10}
  skip_fail: false

- description: Setup Rundeck
  cmd: salt --hard-crash --state-output=mixed --state-verbose=False -C 'I@rundeck:client' state.sls rundeck.client
  node_name: {{ HOSTNAME_CFG01 }}
  retry: {count: 3, delay: 10}
  skip_fail: false

# Elasticsearch (in container, disabled until https://mirantis.jira.com/browse/PROD-15297 is not fixed)
#--------------
#- description: 'Waiting for Elasticsearch to come up in container...'
#  cmd: timeout 30 salt --hard-crash --state-output=mixed --state-verbose=False -C 'I@elasticsearch:client' cmd.run
#      'export CICD_CONTROL_ADDRESS=$(salt-call --out=newline_values_only pillar.get _param:cluster_vip_address);
#       while true; do curl -sf http://${CICD_CONTROL_ADDRESS}:9200/?pretty && break; sleep 2; done'
#  node_name: {{ HOSTNAME_CFG01 }}
#  retry: {count: 3, delay: 10}
#  skip_fail: false
#
#- description: Setup Elasticsearch
#  cmd: salt --hard-crash --state-output=mixed --state-verbose=False -C 'I@elasticsearch:client' state.sls elasticsearch.client
#  node_name: {{ HOSTNAME_CFG01 }}
#  retry: {count: 3, delay: 10}
#  skip_fail: false


# Generate docs
#--------------

- description: Install sphinx (may fail depending on the model)
  cmd: salt --hard-crash --state-output=mixed --state-verbose=False -C 'I@sphinx:server' state.sls sphinx
  node_name: {{ HOSTNAME_CFG01 }}
  retry: {count: 1, delay: 10}
  skip_fail: true

- description: Generate docs (may fail depending on the model)
  cmd: salt-run state.orchestrate sphinx.orch.generate_doc
  node_name: {{ HOSTNAME_CFG01 }}
  retry: {count: 1, delay: 10}
  skip_fail: true

- description: Run salt minion to create cert files for nginx
  cmd: salt --hard-crash --state-output=mixed --state-verbose=False
    -C 'I@nginx:server' state.sls salt.minion
  node_name: {{ HOSTNAME_CFG01 }}
  retry: {count: 1, delay: 10}
  skip_fail: false

- description: Install nginx
  cmd: salt --hard-crash --state-output=mixed --state-verbose=False -C 'I@nginx:server' state.sls nginx
  node_name: {{ HOSTNAME_CFG01 }}
  retry: {count: 3, delay: 10}
  skip_fail: false

# Final checks
#-------------

- description: Check for system services in failed state
  cmd: salt --hard-crash --state-output=mixed --state-verbose=False '*' cmd.run "systemctl --failed | grep -E 'loaded[ \t]+failed' && echo 'Command execution failed'  || true"
  node_name: {{ HOSTNAME_CFG01 }}
  retry: {count: 1, delay: 10}
  skip_fail: false