From dd5d89f7c93e0e94f22268cba897b67e0e5eb5f8 Mon Sep 17 00:00:00 2001 From: Olivier Bourdon Date: Thu, 19 Jan 2017 13:32:11 +0100 Subject: [PATCH] Add alarms for Nova aggregates Change-Id: Ia82d5baf754d2d61c2bfa6d882ace3c8d094eafc Depends-On: I6647600d73991bfbfc7b7c199a7f9b90b9294f68 --- .../templates/alarming.yaml.erb | 35 ++++++++++++++++++- .../templates/clusters.yaml.erb | 6 ++++ doc/user/source/appendix_alarms.rst | 28 +++++++++++++++ 3 files changed, 68 insertions(+), 1 deletion(-) diff --git a/deployment_scripts/puppet/modules/fuel_lma_collector/templates/alarming.yaml.erb b/deployment_scripts/puppet/modules/fuel_lma_collector/templates/alarming.yaml.erb index b9b0206ba..c35e73572 100644 --- a/deployment_scripts/puppet/modules/fuel_lma_collector/templates/alarming.yaml.erb +++ b/deployment_scripts/puppet/modules/fuel_lma_collector/templates/alarming.yaml.erb @@ -1169,7 +1169,7 @@ lma_collector: description: 'There is none memory available for new instances' severity: 'warning' enabled: 'true' - no_data_policy: skip # the metric is only collected from the aggregator node + no_data_policy: skip # the metric is only collected from the aggregator node trigger: rules: - metric: openstack_nova_total_free_ram @@ -1178,6 +1178,34 @@ lma_collector: window: 60 periods: 0 function: max + - name: 'nova-aggregates-free-memory-warning' + description: "The nova aggregates free memory percent is low" + severity: 'warning' + enabled: 'true' + no_data_policy: skip # the metric is only collected from the aggregator node + trigger: + rules: + - metric: openstack_nova_aggregate_free_ram_percent + group_by: [aggregate] + relational_operator: '<' + threshold: 10.0 + window: 60 + periods: 0 + function: min + - name: 'nova-aggregates-free-memory-critical' + description: "The nova aggregates free memory percent is too low" + severity: 'critical' + enabled: 'true' + no_data_policy: skip # the metric is only collected from the aggregator node + trigger: + rules: + - metric: openstack_nova_aggregate_free_ram_percent + group_by: [aggregate] + relational_operator: '<' + threshold: 1.0 + window: 60 + periods: 0 + function: min # Adds alarm on local check for OpenStack services endpoint - name: 'cinder-api-local-endpoint' @@ -3576,6 +3604,11 @@ lma_collector: members: nova-free-memory: alarms: ['total-nova-free-memory-warning'] + nova-aggregates-free-memory: + alerting: enabled + members: + nova-aggregates-free-memory: + alarms: ['nova-aggregates-free-memory-critical', 'nova-aggregates-free-memory-warning'] ceph-mon-cluster: apply_to_node: ceph-mon alerting: enabled diff --git a/deployment_scripts/puppet/modules/fuel_lma_collector/templates/clusters.yaml.erb b/deployment_scripts/puppet/modules/fuel_lma_collector/templates/clusters.yaml.erb index 77f9df3f7..db5faca73 100644 --- a/deployment_scripts/puppet/modules/fuel_lma_collector/templates/clusters.yaml.erb +++ b/deployment_scripts/puppet/modules/fuel_lma_collector/templates/clusters.yaml.erb @@ -634,6 +634,11 @@ lma_collector: group_by: member members: - nova-free-memory + nova-aggregates-free-memory: + policy: highest_severity + group_by: member + members: + - nova-aggregates-free-memory gse_cluster_node: input_message_types: @@ -812,6 +817,7 @@ lma_collector: - libvirt-service - nova-free-vcpu - nova-free-memory + - nova-aggregates-free-memory hints: - neutron-data-plane cinder-control-plane: diff --git a/doc/user/source/appendix_alarms.rst b/doc/user/source/appendix_alarms.rst index 8169649d9..aa84aa8cf 100644 --- a/doc/user/source/appendix_alarms.rst +++ b/doc/user/source/appendix_alarms.rst @@ -1187,6 +1187,34 @@ The following is a list of StackLight built-in alarms:: window: 60 periods: 0 function: max + - name: 'nova-aggregates-free-memory-warning' + description: "The nova aggregates free memory percent is low" + severity: 'warning' + enabled: 'true' + no_data_policy: skip # the metric is only collected from the aggregator node + trigger: + rules: + - metric: openstack_nova_aggregate_free_ram_percent + group_by: [aggregate] + relational_operator: '<' + threshold: 10.0 + window: 60 + periods: 0 + function: min + - name: 'nova-aggregates-free-memory-critical' + description: "The nova aggregates free memory percent is too low" + severity: 'critical' + enabled: 'true' + no_data_policy: skip # the metric is only collected from the aggregator node + trigger: + rules: + - metric: openstack_nova_aggregate_free_ram_percent + group_by: [aggregate] + relational_operator: '<' + threshold: 1.0 + window: 60 + periods: 0 + function: min # Adds alarm on local check for OpenStack services endpoint - name: 'cinder-api-local-endpoint'