diff options
Diffstat (limited to 'roles/openshift_logging_elasticsearch')
5 files changed, 109 insertions, 71 deletions
diff --git a/roles/openshift_logging_elasticsearch/tasks/main.yaml b/roles/openshift_logging_elasticsearch/tasks/main.yaml index ff5ad1045..b731d93a0 100644 --- a/roles/openshift_logging_elasticsearch/tasks/main.yaml +++ b/roles/openshift_logging_elasticsearch/tasks/main.yaml @@ -137,6 +137,16 @@ - "prometheus_out.stderr | length > 0" - "'already exists' not in prometheus_out.stderr" +- set_fact: + _logging_metrics_proxy_passwd: "{{ 16 | lib_utils_oo_random_word | b64encode }}" + +- template: + src: passwd.j2 + dest: "{{mktemp.stdout}}/passwd.yml" + vars: + logging_user_name: "{{ openshift_logging_elasticsearch_prometheus_sa }}" + logging_user_passwd: "{{ _logging_metrics_proxy_passwd }}" + # View role and binding - name: Generate logging-elasticsearch-view-role template: @@ -255,6 +265,8 @@ path: "{{ generated_certs_dir }}/ca.crt" - name: admin.jks path: "{{ generated_certs_dir }}/system.admin.jks" + - name: passwd.yml + path: "{{mktemp.stdout}}/passwd.yml" # services - name: Set logging-{{ es_component }}-cluster service @@ -391,6 +403,7 @@ es_container_security_context: "{{ _es_containers.elasticsearch.securityContext if _es_containers is defined and 'elasticsearch' in _es_containers and 'securityContext' in _es_containers.elasticsearch else None }}" deploy_type: "{{ openshift_logging_elasticsearch_deployment_type }}" es_replicas: 1 + basic_auth_passwd: "{{ _logging_metrics_proxy_passwd | b64decode }}" - name: Set ES dc oc_obj: diff --git a/roles/openshift_logging_elasticsearch/tasks/restart_cluster.yml b/roles/openshift_logging_elasticsearch/tasks/restart_cluster.yml index 6bce13d1d..879459cf6 100644 --- a/roles/openshift_logging_elasticsearch/tasks/restart_cluster.yml +++ b/roles/openshift_logging_elasticsearch/tasks/restart_cluster.yml @@ -1,91 +1,113 @@ --- -# Disable external communication for {{ _cluster_component }} -- name: Disable external communication for logging-{{ _cluster_component }} - oc_service: - state: present - name: "logging-{{ _cluster_component }}" - namespace: "{{ openshift_logging_elasticsearch_namespace }}" - selector: - component: "{{ _cluster_component }}" - provider: openshift - connection: blocked - labels: - logging-infra: 'support' - ports: - - port: 9200 - targetPort: "restapi" - when: - - full_restart_cluster | bool - ## get all pods for the cluster - command: > oc get pod -l component={{ _cluster_component }},provider=openshift -n {{ openshift_logging_elasticsearch_namespace }} -o jsonpath={.items[?(@.status.phase==\"Running\")].metadata.name} register: _cluster_pods -- name: "Disable shard balancing for logging-{{ _cluster_component }} cluster" - command: > - oc exec {{ _cluster_pods.stdout.split(' ')[0] }} -c elasticsearch -n {{ openshift_logging_elasticsearch_namespace }} -- {{ __es_local_curl }} -XPUT 'https://localhost:9200/_cluster/settings' -d '{ "transient": { "cluster.routing.allocation.enable" : "none" } }' - register: _disable_output - changed_when: "'\"acknowledged\":true' in _disable_output.stdout" +### Check for cluster state before making changes -- if its red then we don't want to continue +- name: "Checking current health for {{ _es_node }} cluster" + shell: > + oc exec "{{ _cluster_pods.stdout.split(' ')[0] }}" -c elasticsearch -n "{{ openshift_logging_elasticsearch_namespace }}" -- es_cluster_health + register: _pod_status when: _cluster_pods.stdout_lines | count > 0 -# Flush ES -- name: "Flushing for logging-{{ _cluster_component }} cluster" - command: > - oc exec {{ _cluster_pods.stdout.split(' ')[0] }} -c elasticsearch -n {{ openshift_logging_elasticsearch_namespace }} -- {{ __es_local_curl }} -XPUT 'https://localhost:9200/_flush/synced' - register: _flush_output - changed_when: "'\"acknowledged\":true' in _flush_output.stdout" - when: +- when: + - _pod_status.stdout is defined + - (_pod_status.stdout | from_json)['status'] in ['red'] + block: + - name: Set Logging message to manually restart + run_once: true + set_stats: + data: + installer_phase_logging: + message: "Cluster logging-{{ _cluster_component }} was in a red state and will not be automatically restarted. Please see documentation regarding doing a {{ 'full' if full_restart_cluster | bool else 'rolling'}} cluster restart." + + - debug: msg="Cluster logging-{{ _cluster_component }} was in a red state and will not be automatically restarted. Please see documentation regarding doing a {{ 'full' if full_restart_cluster | bool else 'rolling'}} cluster restart." + +- when: _pod_status.stdout is undefined or (_pod_status.stdout | from_json)['status'] in ['green', 'yellow'] + block: + # Disable external communication for {{ _cluster_component }} + - name: Disable external communication for logging-{{ _cluster_component }} + oc_service: + state: present + name: "logging-{{ _cluster_component }}" + namespace: "{{ openshift_logging_elasticsearch_namespace }}" + selector: + component: "{{ _cluster_component }}" + provider: openshift + connection: blocked + labels: + logging-infra: 'support' + ports: + - port: 9200 + targetPort: "restapi" + when: + - full_restart_cluster | bool + + - name: "Disable shard balancing for logging-{{ _cluster_component }} cluster" + command: > + oc exec {{ _cluster_pods.stdout.split(' ')[0] }} -c elasticsearch -n {{ openshift_logging_elasticsearch_namespace }} -- {{ __es_local_curl }} -XPUT 'https://localhost:9200/_cluster/settings' -d '{ "transient": { "cluster.routing.allocation.enable" : "none" } }' + register: _disable_output + changed_when: "'\"acknowledged\":true' in _disable_output.stdout" + when: _cluster_pods.stdout_lines | count > 0 + + # Flush ES + - name: "Flushing for logging-{{ _cluster_component }} cluster" + command: > + oc exec {{ _cluster_pods.stdout.split(' ')[0] }} -c elasticsearch -n {{ openshift_logging_elasticsearch_namespace }} -- {{ __es_local_curl }} -XPUT 'https://localhost:9200/_flush/synced' + register: _flush_output + changed_when: "'\"acknowledged\":true' in _flush_output.stdout" + when: - _cluster_pods.stdout_lines | count > 0 - full_restart_cluster | bool -- command: > - oc get dc -l component={{ _cluster_component }},provider=openshift -n {{ openshift_logging_elasticsearch_namespace }} -o jsonpath={.items[*].metadata.name} - register: _cluster_dcs + - command: > + oc get dc -l component={{ _cluster_component }},provider=openshift -n {{ openshift_logging_elasticsearch_namespace }} -o jsonpath={.items[*].metadata.name} + register: _cluster_dcs -## restart all dcs for full restart -- name: "Restart ES node {{ _es_node }}" - include_tasks: restart_es_node.yml - with_items: "{{ _cluster_dcs }}" - loop_control: - loop_var: _es_node - when: + ## restart all dcs for full restart + - name: "Restart ES node {{ _es_node }}" + include_tasks: restart_es_node.yml + with_items: "{{ _cluster_dcs }}" + loop_control: + loop_var: _es_node + when: - full_restart_cluster | bool -## restart the node if it's dc is in the list of nodes to restart? -- name: "Restart ES node {{ _es_node }}" - include_tasks: restart_es_node.yml - with_items: "{{ _restart_logging_nodes }}" - loop_control: - loop_var: _es_node - when: + ## restart the node if it's dc is in the list of nodes to restart? + - name: "Restart ES node {{ _es_node }}" + include_tasks: restart_es_node.yml + with_items: "{{ _restart_logging_nodes }}" + loop_control: + loop_var: _es_node + when: - not full_restart_cluster | bool - _es_node in _cluster_dcs.stdout -## we may need a new first pod to run against -- fetch them all again -- command: > - oc get pod -l component={{ _cluster_component }},provider=openshift -n {{ openshift_logging_elasticsearch_namespace }} -o jsonpath={.items[?(@.status.phase==\"Running\")].metadata.name} - register: _cluster_pods + ## we may need a new first pod to run against -- fetch them all again + - command: > + oc get pod -l component={{ _cluster_component }},provider=openshift -n {{ openshift_logging_elasticsearch_namespace }} -o jsonpath={.items[?(@.status.phase==\"Running\")].metadata.name} + register: _cluster_pods -- name: "Enable shard balancing for logging-{{ _cluster_component }} cluster" - command: > - oc exec {{ _cluster_pods.stdout.split(' ')[0] }} -c elasticsearch -n {{ openshift_logging_elasticsearch_namespace }} -- {{ __es_local_curl }} -XPUT 'https://localhost:9200/_cluster/settings' -d '{ "transient": { "cluster.routing.allocation.enable" : "all" } }' - register: _enable_output - changed_when: "'\"acknowledged\":true' in _enable_output.stdout" + - name: "Enable shard balancing for logging-{{ _cluster_component }} cluster" + command: > + oc exec {{ _cluster_pods.stdout.split(' ')[0] }} -c elasticsearch -n {{ openshift_logging_elasticsearch_namespace }} -- {{ __es_local_curl }} -XPUT 'https://localhost:9200/_cluster/settings' -d '{ "transient": { "cluster.routing.allocation.enable" : "all" } }' + register: _enable_output + changed_when: "'\"acknowledged\":true' in _enable_output.stdout" -# Reenable external communication for {{ _cluster_component }} -- name: Reenable external communication for logging-{{ _cluster_component }} - oc_service: - state: present - name: "logging-{{ _cluster_component }}" - namespace: "{{ openshift_logging_elasticsearch_namespace }}" - selector: - component: "{{ _cluster_component }}" - provider: openshift - labels: - logging-infra: 'support' - ports: + # Reenable external communication for {{ _cluster_component }} + - name: Reenable external communication for logging-{{ _cluster_component }} + oc_service: + state: present + name: "logging-{{ _cluster_component }}" + namespace: "{{ openshift_logging_elasticsearch_namespace }}" + selector: + component: "{{ _cluster_component }}" + provider: openshift + labels: + logging-infra: 'support' + ports: - port: 9200 targetPort: "restapi" - when: + when: - full_restart_cluster | bool diff --git a/roles/openshift_logging_elasticsearch/tasks/restart_es_node.yml b/roles/openshift_logging_elasticsearch/tasks/restart_es_node.yml index 6d0df40c8..fe15e40fd 100644 --- a/roles/openshift_logging_elasticsearch/tasks/restart_es_node.yml +++ b/roles/openshift_logging_elasticsearch/tasks/restart_es_node.yml @@ -26,12 +26,12 @@ - name: "Waiting for ES to be ready for {{ _es_node }}" shell: > - oc exec "{{ _pod }}" -c elasticsearch -n "{{ openshift_logging_elasticsearch_namespace }}" -- {{ __es_local_curl }} https://localhost:9200/_cat/health | cut -d' ' -f4 + oc exec "{{ _pod }}" -c elasticsearch -n "{{ openshift_logging_elasticsearch_namespace }}" -- es_cluster_health with_items: "{{ _pods.stdout.split(' ') }}" loop_control: loop_var: _pod register: _pod_status - until: _pod_status.stdout in ['green', 'yellow'] + until: (_pod_status.stdout | from_json)['status'] in ['green', 'yellow'] retries: 60 delay: 5 changed_when: false diff --git a/roles/openshift_logging_elasticsearch/templates/es.j2 b/roles/openshift_logging_elasticsearch/templates/es.j2 index 4b189f255..b1d6a4489 100644 --- a/roles/openshift_logging_elasticsearch/templates/es.j2 +++ b/roles/openshift_logging_elasticsearch/templates/es.j2 @@ -51,6 +51,7 @@ spec: - -client-id={{openshift_logging_elasticsearch_prometheus_sa}} - -client-secret-file=/var/run/secrets/kubernetes.io/serviceaccount/token - -cookie-secret={{ 16 | lib_utils_oo_random_word | b64encode }} + - -basic-auth-password={{ basic_auth_passwd }} - -upstream=https://localhost:9200 - '-openshift-sar={"namespace": "{{ openshift_logging_elasticsearch_namespace}}", "verb": "view", "resource": "prometheus", "group": "metrics.openshift.io"}' - '-openshift-delegate-urls={"/": {"resource": "prometheus", "verb": "view", "group": "metrics.openshift.io", "namespace": "{{ openshift_logging_elasticsearch_namespace}}"}}' diff --git a/roles/openshift_logging_elasticsearch/templates/passwd.j2 b/roles/openshift_logging_elasticsearch/templates/passwd.j2 new file mode 100644 index 000000000..a22151eef --- /dev/null +++ b/roles/openshift_logging_elasticsearch/templates/passwd.j2 @@ -0,0 +1,2 @@ +"{{logging_user_name}}": + passwd: "{{logging_user_passwd}}" |