Merge "Un-hardcoding stricthostkeychecking for nova user" into release/2019.2.0
diff --git a/_states/novav21.py b/_states/novav21.py
index b9f9dd0..7b6452b 100644
--- a/_states/novav21.py
+++ b/_states/novav21.py
@@ -361,6 +361,43 @@
return ret
+def instances_mapped_to_cell(name, timeout=60, runas='nova'):
+ """Ensure that all instances in the cell are mapped
+
+ :param name: cell name.
+ :param timeout: amount of time in seconds mapping process should finish in.
+ :param runas: username to run the shell commands under.
+ """
+ test = __opts__.get('test', False)
+ cell_uuid = __salt__['cmd.shell'](
+ "nova-manage cell_v2 list_cells 2>/dev/null | "
+ "awk '/%s/ {print $4}'" % name, runas=runas)
+ result = {'name': name, 'changes': {}, 'result': False}
+ if not cell_uuid:
+ result['comment'] = (
+ 'Failed to map all instances in cell {0}, it does not exist'
+ .format(name))
+ return result
+ start_time = time.time()
+ if not test:
+ while True:
+ rc = __salt__['cmd.retcode'](
+ 'nova-manage cell_v2 map_instances --cell_uuid %s' % cell_uuid,
+ runas=runas)
+ if rc == 0 or time.time() - start_time > timeout:
+ break
+ if rc != 0:
+ result['comment'] = (
+ 'Failed to map all instances in cell {0} in {1} seconds'
+ .format(name, timeout))
+ return result
+ result['comment'] = 'All instances mapped in cell {0}'.format(name)
+ if test:
+ result['comment'] = 'TEST: {}'.format(result['comment'])
+ result['result'] = True
+ return result
+
+
def _db_version_update(db, version, human_readable_resource_name):
existing_version = __salt__['cmd.shell'](
'nova-manage %s version 2>/dev/null' % db)
diff --git a/nova/map.jinja b/nova/map.jinja
index 370f517..d9fea7b 100644
--- a/nova/map.jinja
+++ b/nova/map.jinja
@@ -295,20 +295,6 @@
'warn': '15%',
'crit': '5%',
},
- 'error_log_rate': {
- 'warn': 0.2,
- },
- 'services_failed_warning_threshold_percent': 0.3,
- 'services_failed_critical_threshold_percent': 0.6,
- 'computes_failed_warning_threshold_percent': 0.25,
- 'computes_failed_critical_threshold_percent': 0.5,
- 'cpu_minor_threshold': 0.85,
- 'cpu_major_threshold': 0.95,
- 'ram_major_threshold': 0.85,
- 'ram_critical_threshold': 0.95,
- 'disk_major_threshold': 0.85,
- 'disk_critical_threshold': 0.95,
- 'endpoint_failed_major_threshold': 0.5,
},
}, grain='os_family', merge=salt['pillar.get']('nova:monitoring')) %}
diff --git a/nova/meta/prometheus.yml b/nova/meta/prometheus.yml
index 1950e9b..f3e12ee 100644
--- a/nova/meta/prometheus.yml
+++ b/nova/meta/prometheus.yml
@@ -1,4 +1,4 @@
-{% from "nova/map.jinja" import controller, compute, monitoring with context %}
+{% from "nova/map.jinja" import controller, compute with context %}
{%- set is_controller = controller.get('enabled', False) %}
{%- set is_compute = compute.get('enabled', False) %}
@@ -29,11 +29,6 @@
server:
alert:
{%- if is_controller %}
-{%- set minor_threshold = monitoring.services_failed_warning_threshold_percent|float %}
-{%- set major_threshold = monitoring.services_failed_critical_threshold_percent|float %}
-{%- set minor_compute_threshold = monitoring.computes_failed_warning_threshold_percent|float %}
-{%- set major_compute_threshold = monitoring.computes_failed_critical_threshold_percent|float %}
-{%- set major_endpoint_threshold = monitoring.endpoint_failed_major_threshold|float %}
{% raw %}
NovaApiOutage:
if: >-
@@ -66,18 +61,17 @@
summary: "nova-api endpoint is not accessible"
description: >-
The nova-api endpoint on the {{ $labels.host }} node is not accessible for 2 minutes.
-{%- endraw %}
NovaApiEndpointsDownMajor:
if: >-
- count(http_response_status{name=~"nova-api"} == 0) >= count(http_response_status{name=~"nova-api"}) * {{ major_endpoint_threshold }}
+ count(http_response_status{name=~"nova-api"} == 0) >= count(http_response_status{name=~"nova-api"}) * 0.6
for: 2m
labels:
severity: major
service: nova
annotations:
- summary: "{{major_endpoint_threshold * 100}}% of nova-api endpoints are not accessible"
+ summary: "60% of nova-api endpoints are not accessible"
description: >-
- {% raw %}{{ $value }} nova-api endpoints (>= {% endraw %} {{major_endpoint_threshold * 100}}{% raw %}%) are not accessible for 2 minutes.
+ More than 60% of nova-api endpoints are not accessible for 2 minutes.
NovaApiEndpointsOutage:
if: >-
count(http_response_status{name=~"nova-api"} == 0) == count(http_response_status{name=~"nova-api"})
@@ -99,47 +93,46 @@
summary: "{{ $labels.binary }} service is down"
description: >-
The {{ $labels.binary }} service on the {{ $labels.hostname }} node is down.
-{%- endraw %}
NovaServicesDownMinor:
if: >-
- count(openstack_nova_service_state{binary!~"nova-compute"} == 0) by (binary) >= on (binary) count(openstack_nova_service_state{binary!~"nova-compute"}) by (binary) * {{minor_threshold}}
+ count(openstack_nova_service_state{binary!~"nova-compute"} == 0) by (binary) >= on (binary) count(openstack_nova_service_state{binary!~"nova-compute"}) by (binary) * 0.3
labels:
severity: minor
service: nova
annotations:
- summary: "{{minor_threshold * 100}}%{%- raw %} of {{ $labels.binary }} services are down"
+ summary: "30% of {{ $labels.binary }} services are down"
description: >-
- {{ $value }} {{ $labels.binary }} services (>= {%- endraw %} {{minor_threshold * 100}}%) are down.
+ More than 30% {{ $labels.binary }} services are down.
NovaComputeServicesDownMinor:
if: >-
- count(openstack_nova_service_state{binary="nova-compute"} == 0) >= count(openstack_nova_service_state{binary="nova-compute"}) * {{minor_compute_threshold}}
+ count(openstack_nova_service_state{binary="nova-compute"} == 0) >= count(openstack_nova_service_state{binary="nova-compute"}) * 0.25
labels:
severity: minor
service: nova
annotations:
- summary: "{{minor_compute_threshold * 100}}%{%- raw %} of nova-compute services are down"
+ summary: "More than 25% of nova-compute services are down"
description: >-
- {{ $value }} nova-compute services (>= {%- endraw %} {{minor_compute_threshold * 100}}%) are down.
+ More than 25% of nova-compute services are down.
NovaServicesDownMajor:
if: >-
- count(openstack_nova_service_state{binary!~"nova-compute"} == 0) by (binary) >= on (binary) count(openstack_nova_service_state{binary!~"nova-compute"}) by (binary) * {{major_threshold}}
+ count(openstack_nova_service_state{binary!~"nova-compute"} == 0) by (binary) >= on (binary) count(openstack_nova_service_state{binary!~"nova-compute"}) by (binary) * 0.6
labels:
severity: major
service: nova
annotations:
- summary: "{{major_threshold * 100}}%{%- raw %} of {{ $labels.binary }} services are down"
+ summary: "More than 60% of {{ $labels.binary }} services are down"
description: >-
- {{ $value }} {{ $labels.binary }} services (>= {%- endraw %} {{major_threshold * 100}}%) are down.
+ More than 60% of {{ $labels.binary }} services are down.
NovaComputeServicesDownMajor:
if: >-
- count(openstack_nova_service_state{binary="nova-compute"} == 0) >= count(openstack_nova_service_state{binary="nova-compute"}) * {{major_compute_threshold}}
+ count(openstack_nova_service_state{binary="nova-compute"} == 0) >= count(openstack_nova_service_state{binary="nova-compute"}) * 0.5
labels:
severity: major
service: nova
annotations:
- summary: "{{major_compute_threshold * 100}}%{%- raw %} of nova-compute services are down"
+ summary: "More than 50% of nova-compute services are down"
description: >-
- {{ $value }} nova-compute services (>= {%- endraw %} {{major_compute_threshold * 100}}{%- raw %}%) are down.
+ More than 50% of nova-compute services are down.
NovaServiceOutage:
if: >-
count(openstack_nova_service_state == 0) by (binary) == on (binary) count(openstack_nova_service_state) by (binary)
@@ -151,168 +144,17 @@
description: >-
All {{ $labels.binary }} services are down.
{%- endraw %}
-{%- set cpu_minor_threshold = monitoring.cpu_minor_threshold|float %}
-{%- set cpu_major_threshold = monitoring.cpu_major_threshold|float %}
-{%- set ram_major_threshold = monitoring.ram_major_threshold|float %}
-{%- set ram_critical_threshold = monitoring.ram_critical_threshold|float %}
-{%- set disk_major_threshold = monitoring.disk_major_threshold|float %}
-{%- set disk_critical_threshold = monitoring.disk_critical_threshold|float %}
- NovaHypervisorVCPUsFullMinor:
- if: >-
- label_replace(system_load15, "hostname", "$1", "host", "(.*)") > on (hostname) openstack_nova_vcpus * {{ cpu_minor_threshold }}
- labels:
- severity: minor
- service: nova
- annotations:
- summary: "{{ cpu_minor_threshold * 100 }}% of hypervisor VCPUs are used"
- description: "{% raw %}{{ $value }} VCPUs on the {{ $labels.hostname }} node (> {% endraw %} {{ cpu_minor_threshold * 100 }}%) are used."
- NovaHypervisorVCPUsFullMajor:
- if: >-
- label_replace(system_load15, "hostname", "$1", "host", "(.*)") > on (hostname) openstack_nova_vcpus * {{ cpu_major_threshold }}
- labels:
- severity: major
- service: nova
- annotations:
- summary: "{{ cpu_major_threshold * 100 }}% of hypervisor VCPUs are used"
- description: "{% raw %}{{ $value }} VCPUs on the {{ $labels.hostname }} node (> {% endraw %} {{ cpu_major_threshold * 100 }}%) are used."
- NovaHypervisorMemoryFullMajor:
- if: >-
- openstack_nova_used_ram > openstack_nova_ram * {{ ram_major_threshold }}
- labels:
- severity: major
- service: nova
- annotations:
- summary: "{{ ram_major_threshold * 100 }}% of hypervisor RAM is used"
- description: "{% raw %}{{ $value }}MB of RAM on the {{ $labels.hostname }} node (> {% endraw %} {{ ram_major_threshold * 100 }}%) is used."
- NovaHypervisorMemoryFullCritical:
- if: >-
- openstack_nova_used_ram > openstack_nova_ram * {{ ram_critical_threshold }}
- labels:
- severity: critical
- service: nova
- annotations:
- summary: "{{ ram_critical_threshold * 100 }}% of hypervisor RAM is used"
- description: "{% raw %}{{ $value }}MB of RAM on the {{ $labels.hostname }} node (> {% endraw %} {{ ram_critical_threshold * 100 }}%) is used."
- NovaHypervisorDiskFullMajor:
- if: >-
- openstack_nova_used_disk > openstack_nova_disk * {{ disk_major_threshold }}
- labels:
- severity: major
- service: nova
- annotations:
- summary: "{{ disk_major_threshold * 100 }}% of hypervisor disk space is used"
- description: "{% raw %}{{ $value }}GB of disk space on the {{ $labels.hostname }} node (> {% endraw %} {{ disk_major_threshold * 100 }}%) is used."
- NovaHypervisorDiskFullCritical:
- if: >-
- openstack_nova_used_disk > openstack_nova_disk * {{ disk_critical_threshold }}
- labels:
- severity: critical
- service: nova
- annotations:
- summary: "{{ disk_critical_threshold * 100 }}% of hypervisor disk space is used"
- description: "{% raw %}{{ $value }}GB of disk space on the {{ $labels.hostname }} node (> {% endraw %} {{ disk_critical_threshold * 100 }}%) is used."
- NovaAggregateMemoryFullMajor:
- if: >-
- openstack_nova_aggregate_used_ram > openstack_nova_aggregate_ram * {{ ram_major_threshold }}
- labels:
- severity: major
- service: nova
- annotations:
- summary: "{{ ram_major_threshold * 100 }}% of aggregate RAM is used"
- description: "{% raw %}{{ $value }}MB of RAM on the {{ $labels.aggregate }} aggregate (> {% endraw %} {{ ram_major_threshold * 100 }}%) is used."
- NovaAggregateMemoryFullCritical:
- if: >-
- openstack_nova_aggregate_used_ram > openstack_nova_aggregate_ram * {{ ram_critical_threshold }}
- labels:
- severity: critical
- service: nova
- annotations:
- summary: "{{ ram_critical_threshold * 100 }}% of aggregate RAM is used"
- description: "{% raw %}{{ $value }}MB of RAM on the {{ $labels.aggregate }} aggregate (> {% endraw %} {{ ram_critical_threshold * 100 }}%) is used."
- NovaAggregateDiskFullMajor:
- if: >-
- openstack_nova_aggregate_used_disk > openstack_nova_aggregate_disk * {{ disk_major_threshold }}
- labels:
- severity: major
- service: nova
- annotations:
- summary: "{{ disk_major_threshold * 100 }}% of aggregate disk space is used"
- description: "{% raw %}{{ $value }}GB of disk space on the {{ $labels.aggregate }} aggregate (> {% endraw %} {{ disk_major_threshold * 100 }}%) is used."
- NovaAggregateDiskFullCritical:
- if: >-
- openstack_nova_aggregate_used_disk > openstack_nova_aggregate_disk * {{ disk_critical_threshold }}
- labels:
- severity: critical
- service: nova
- annotations:
- summary: "{{ disk_critical_threshold * 100 }}% of aggregate disk space is used"
- description: "{% raw %}{{ $value }}GB of disk space on the {{ $labels.aggregate }} aggregate (> {% endraw %} {{ disk_critical_threshold * 100 }}%) is used."
- NovaTotalVCPUsFullMinor:
- if: >-
- sum(label_replace(system_load15, "hostname", "$1", "host", "(.*)") and on (hostname) openstack_nova_vcpus) > max(sum(openstack_nova_vcpus) by (instance)) * {{ cpu_minor_threshold }}
- labels:
- severity: minor
- service: nova
- annotations:
- summary: "{{ cpu_minor_threshold * 100 }}% of cloud VCPUs are used"
- description: "{% raw %}{{ $value }} VCPUs in the cloud (> {% endraw %} {{ cpu_minor_threshold * 100 }}%) are used."
- NovaTotalVCPUsFullMajor:
- if: >-
- sum(label_replace(system_load15, "hostname", "$1", "host", "(.*)") and on (hostname) openstack_nova_vcpus) > max(sum(openstack_nova_vcpus) by (instance)) * {{ cpu_major_threshold }}
- labels:
- severity: major
- service: nova
- annotations:
- summary: "{{ cpu_major_threshold * 100 }}% of cloud VCPUs are used"
- description: "{% raw %}{{ $value }} VCPUs in the cloud (> {% endraw %} {{ cpu_major_threshold * 100 }}%) are used."
- NovaTotalMemoryFullMajor:
- if: >-
- openstack_nova_total_used_ram > openstack_nova_total_ram * {{ ram_major_threshold }}
- labels:
- severity: major
- service: nova
- annotations:
- summary: "{{ ram_major_threshold * 100 }}% of cloud RAM is used"
- description: "{% raw %}{{ $value }}MB of RAM in the cloud (> {% endraw %} {{ ram_major_threshold * 100 }}%) is used."
- NovaTotalMemoryFullCritical:
- if: >-
- openstack_nova_total_used_ram > openstack_nova_total_ram * {{ ram_critical_threshold }}
- labels:
- severity: critical
- service: nova
- annotations:
- summary: "{{ ram_critical_threshold * 100 }}% of cloud RAM is used"
- description: "{% raw %}{{ $value }}MB of RAM in the cloud (> {% endraw %} {{ ram_critical_threshold * 100 }}%) is used."
- NovaTotalDiskFullMajor:
- if: >-
- openstack_nova_total_used_disk > openstack_nova_total_disk * {{ disk_major_threshold }}
- labels:
- severity: major
- service: nova
- annotations:
- summary: "{{ disk_major_threshold * 100 }}% of cloud disk space is used"
- description: "{% raw %}{{ $value }}GB of disk space in the cloud (> {% endraw %} {{ disk_major_threshold * 100 }}%) is used."
- NovaTotalDiskFullCritical:
- if: >-
- openstack_nova_total_used_disk > openstack_nova_total_disk * {{ disk_critical_threshold }}
- labels:
- severity: critical
- service: nova
- annotations:
- summary: "{{ disk_critical_threshold * 100 }}% of cloud disk space is used"
- description: "{% raw %}{{ $value }}GB of disk space in the cloud (> {% endraw %} {{ disk_critical_threshold * 100 }}%) is used."
{%- endif %}
- NovaErrorLogsTooHigh:
- {%- set log_threshold = monitoring.error_log_rate.warn|float %}
- if: >-
- sum(rate(log_messages{service="nova",level=~"(?i:(error|emergency|fatal))"}[5m])) without (level) > {{ log_threshold }}
{%- raw %}
+ NovaErrorLogsTooHigh:
+ if: >-
+ sum(rate(log_messages{service="nova",level=~"(?i:(error|emergency|fatal))"}[5m])) without (level) > 0.2
labels:
severity: warning
service: nova
annotations:
summary: "High number of errors in Nova logs"
- description: "The average per-second rate of errors in Nova logs on the {{ $labels.host }} node is {{ $value }} (as measured over the last 5 minutes)."
+ description: "The average rate of errors in Nova logs on the {{ $labels.host }} node is more than 0.2 error messages per second (as measured over the last 5 minutes)."
{%- endraw %}
{%- if is_compute and exporters is defined and compute.get('compute_driver', 'libvirt.LibvirtDriver') == 'libvirt.LibvirtDriver'%}
{%- raw %}
diff --git a/nova/upgrade/pre/init.sls b/nova/upgrade/pre/init.sls
index 9ed049a..0c5834e 100644
--- a/nova/upgrade/pre/init.sls
+++ b/nova/upgrade/pre/init.sls
@@ -16,7 +16,7 @@
/etc/nova/nova.conf:
file.managed:
- name: /etc/nova/nova.conf
- - source: salt://nova/files/{{ _data.version }}/nova-{{ type }}.conf.{{ grains.os_family }}
+ - source: salt://nova/files/{{ upgrade.old_release }}/nova-{{ type }}.conf.{{ grains.os_family }}
- template: jinja
{%- if controller.get('enabled') %}