Add KVM feature toggle to datasources, knowledges, kpis, and pipelines

helm/bundles/cortex-nova/templates/datasources.yaml

@@ -73,35 +73,6 @@ spec:
 ---
 apiVersion: cortex.cloud/v1alpha1
 kind: Datasource
-metadata:
-  name: kvm-libvirt-domain-steal-pct
-spec:
-  schedulingDomain: nova
-  databaseSecretRef:
-    name: cortex-nova-postgres
-    namespace: {{ .Release.Namespace }}
-  {{- if .Values.prometheus.sso.enabled }}
-  ssoSecretRef:
-    name: cortex-nova-prometheus-sso
-    namespace: {{ .Release.Namespace }}
-  {{- end }}
-  type: prometheus
-  prometheus:
-    secretRef:
-      name: cortex-nova-prometheus
-      namespace: {{ .Release.Namespace }}
-    alias: kvm_libvirt_domain_steal_pct
-    # This metric is exported by https://github.com/cobaltcore-dev/kvm-monitoring
-    query: |
-      max by (domain) (rate(kvm_domain_libvirt_vcpu_delay_nanoseconds[5m])) / 1e9 * 100
-    type: kvm_libvirt_domain_metric
-    # It's ok to only look at a short time period here.
-    timeRange: "1200s" # 20 minutes
-    interval: "300s" # 5 minutes
-    resolution: "60s" # 1 minute
----
-apiVersion: cortex.cloud/v1alpha1
-kind: Datasource
 metadata:
   name: nova-servers
 spec:

helm/bundles/cortex-nova/templates/datasources_kvm.yaml

@@ -0,0 +1,31 @@
+{{- if .Values.kvm.enabled }}
+---
+apiVersion: cortex.cloud/v1alpha1
+kind: Datasource
+metadata:
+  name: kvm-libvirt-domain-steal-pct
+spec:
+  schedulingDomain: nova
+  databaseSecretRef:
+    name: cortex-nova-postgres
+    namespace: {{ .Release.Namespace }}
+  {{- if .Values.prometheus.sso.enabled }}
+  ssoSecretRef:
+    name: cortex-nova-prometheus-sso
+    namespace: {{ .Release.Namespace }}
+  {{- end }}
+  type: prometheus
+  prometheus:
+    secretRef:
+      name: cortex-nova-prometheus
+      namespace: {{ .Release.Namespace }}
+    alias: kvm_libvirt_domain_steal_pct
+    # This metric is exported by https://github.com/cobaltcore-dev/kvm-monitoring
+    query: |
+      max by (domain) (rate(kvm_domain_libvirt_vcpu_delay_nanoseconds[5m])) / 1e9 * 100
+    type: kvm_libvirt_domain_metric
+    # It's ok to only look at a short time period here.
+    timeRange: "1200s" # 20 minutes
+    interval: "300s" # 5 minutes
+    resolution: "60s" # 1 minute
+{{- end }}

helm/bundles/cortex-nova/templates/knowledges.yaml

@@ -70,22 +70,8 @@ spec:
       - name: vmware-resolved-hostsystems
     datasources:
       - name: vrops-hostsystem-cpu-contention-short-term-percentage
----
-apiVersion: cortex.cloud/v1alpha1
-kind: Knowledge
-metadata:
-  name: kvm-libvirt-domain-cpu-steal-pct
-spec:
-  schedulingDomain: nova
-  extractor:
-    name: kvm_libvirt_domain_cpu_steal_pct_extractor
-  description: |
-    This knowledge identifies KVM libvirt domains that are experiencing CPU
-    steal, i.e. cpu contention or cpu wait.
-  dependencies:
-    datasources:
-      - name: kvm-libvirt-domain-steal-pct
-      - name: nova-servers
+
+
 ---
 apiVersion: cortex.cloud/v1alpha1
 kind: Knowledge

helm/bundles/cortex-nova/templates/knowledges_kvm.yaml

@@ -0,0 +1,18 @@
+{{- if .Values.kvm.enabled }}
+---
+apiVersion: cortex.cloud/v1alpha1
+kind: Knowledge
+metadata:
+  name: kvm-libvirt-domain-cpu-steal-pct
+spec:
+  schedulingDomain: nova
+  extractor:
+    name: kvm_libvirt_domain_cpu_steal_pct_extractor
+  description: |
+    This knowledge identifies KVM libvirt domains that are experiencing CPU
+    steal, i.e. cpu contention or cpu wait.
+  dependencies:
+    datasources:
+      - name: kvm-libvirt-domain-steal-pct
+      - name: nova-servers
+{{- end }}

helm/bundles/cortex-nova/templates/kpis.yaml

@@ -110,20 +110,6 @@ spec:
 ---
 apiVersion: cortex.cloud/v1alpha1
 kind: KPI
-metadata:
-  name: kvm-host-capacity
-spec:
-  schedulingDomain: nova
-  impl: kvm_host_capacity_kpi
-  dependencies:
-    knowledges:
-      - name: host-details
-      - name: host-utilization
-  description: |
-    This KPI tracks the total, utilized, reserved and failover capacity of KVM hosts.
----
-apiVersion: cortex.cloud/v1alpha1
-kind: KPI
 metadata:
   name: cortex-nova-datasource-state
 spec:

helm/bundles/cortex-nova/templates/kpis_kvm.yaml

@@ -0,0 +1,16 @@
+{{- if .Values.kvm.enabled }}
+---
+apiVersion: cortex.cloud/v1alpha1
+kind: KPI
+metadata:
+  name: kvm-host-capacity
+spec:
+  schedulingDomain: nova
+  impl: kvm_host_capacity_kpi
+  dependencies:
+    knowledges:
+      - name: host-details
+      - name: host-utilization
+  description: |
+    This KPI tracks the total, utilized, reserved and failover capacity of KVM hosts.
+{{- end }}

helm/bundles/cortex-nova/templates/pipelines.yaml

@@ -83,167 +83,3 @@ spec:
         maxCPUContentionActivationLowerBound: 0.0
         maxCPUContentionActivationUpperBound: -0.25
       mandatory: false
----
-apiVersion: cortex.cloud/v1alpha1
-kind: Pipeline
-metadata:
-  name: nova-external-scheduler-kvm
-spec:
-  schedulingDomain: nova
-  description: |
-    Nova provides virtual machine placement on compute hosts for OpenStack.
-    After applying its own filtering and weighing logic, it delegates to cortex
-    for additional filtering and weighing via this external scheduler pipeline.
-    Cortex returns a ranked list of hosts back to nova for final selection.
-    This is the pipeline used for KVM hypervisors (qemu and cloud-hypervisor).
-  type: filter-weigher
-  {{- if $createDecisions }}
-  createDecisions: true
-  {{- end }}
-  steps: []
----
-apiVersion: cortex.cloud/v1alpha1
-kind: Pipeline
-metadata:
-  name: nova-external-scheduler-kvm-all-filters-enabled
-spec:
-  schedulingDomain: nova
-  description: |
-    This pipeline can be used to place reservations the same way nova would place
-    its virtual machines. It uses the same filtering steps as implemented in the
-    nova service, ensuring a valid placement of the reservation. It then leverages
-    cortex's weighing steps to provide an optimized host selection for the reservation.
-    This is the pipeline used for KVM hypervisors (qemu and cloud-hypervisor).
-  type: filter-weigher
-  {{- if $createDecisions }}
-  createDecisions: true
-  {{- end }}
-  steps:
-    - type: filter
-      impl: filter_host_instructions
-      description: |
-        This step will consider the `ignore_hosts` and `force_hosts` instructions
-        from the nova scheduler request spec to filter out or exclusively allow
-        certain hosts.
-      knowledges: []
-    - type: filter
-      impl: filter_has_enough_capacity
-      description: |
-        This step will filter out hosts that do not have enough available capacity
-        to host the requested flavor. If enabled, this step will subtract the
-        current reservations residing on this host from the available capacity.
-      opts:
-        # If reserved space should be locked even for matching requests.
-        # For the reservations pipeline, we don't want to unlock
-        # reserved space, to avoid reservations for the same project
-        # and flavor to overlap.
-        lockReserved: true
-    - type: filter
-      impl: filter_has_requested_traits
-      description: |
-        This step filters hosts that do not have the requested traits given by the
-        nova flavor extra spec: "trait:<trait>": "forbidden" means the host must
-        not have the specified trait. "trait:<trait>": "required" means the host
-        must have the specified trait.
-    - type: filter
-      impl: filter_has_accelerators
-      description: |
-        This step will filter out hosts without the trait `COMPUTE_ACCELERATORS` if
-        the nova flavor extra specs request accelerators via "accel:device_profile".
-    - type: filter
-      impl: filter_correct_az
-      description: |
-        This step will filter out hosts whose aggregate information indicates they
-        are not placed in the requested availability zone.
-    - type: filter
-      impl: filter_status_conditions
-      description: |
-        This step will filter out hosts for which the hypervisor status conditions
-        do not meet the expected values, for example, that the hypervisor is ready
-        and not disabled.
-    - type: filter
-      impl: filter_maintenance
-      description: |
-        This step will filter out hosts that are currently in maintenance mode that
-        prevents scheduling, for example, manual maintenance or termination.
-    - type: filter
-      impl: filter_external_customer
-      description: |
-        This step prefix-matches the domain name for external customer domains and
-        filters out hosts that are not intended for external customers. It considers
-        the `CUSTOM_EXTERNAL_CUSTOMER_SUPPORTED` trait on hosts as well as the
-        `domain_name` scheduler hint from the nova request spec.
-      opts:
-        domainNamePrefixes: ["iaas-"]
-    - type: filter
-      impl: filter_packed_virtqueue
-      description: |
-        If the flavor extra specs contain the `hw:virtio_packed_ring` key, or the
-        image properties contain the `hw_virtio_packed_ring` key, this step will
-        filter out hosts that do not have the `COMPUTE_NET_VIRTIO_PACKED` trait.
-    - type: filter
-      impl: filter_allowed_projects
-      description: |
-        This step filters hosts based on allowed projects defined in the
-        hypervisor resource. Note that hosts allowing all projects are still
-        accessible and will not be filtered out. In this way some hypervisors
-        are made accessible to some projects only.
-    - type: filter
-      impl: filter_capabilities
-      description: |
-        This step will filter out hosts that do not meet the compute capabilities
-        requested by the nova flavor extra specs, like `{"arch": "x86_64",
-        "maxphysaddr:bits": 46, ...}`.
-
-        Note: currently, advanced boolean/numeric operators for the capabilities
-        like `>`, `!`, ... are not supported because they are not used by any of our
-        flavors in production.
-    - type: filter
-      impl: filter_instance_group_affinity
-      description: |
-        This step selects hosts in the instance group specified in the nova
-        scheduler request spec.
-    - type: filter
-      impl: filter_instance_group_anti_affinity
-      description: |
-        This step selects hosts not in the instance group specified in the nova
-        scheduler request spec, but only until the max_server_per_host limit is
-        reached (default = 1).
-    - type: filter
-      impl: filter_live_migratable
-      description: |
-        This step ensures that the target host of a live migration can accept
-        the migrating VM, by checking cpu architecture, cpu features, emulated
-        devices, and cpu modes.
-    - type: filter
-      impl: filter_requested_destination
-      description: |
-        This step filters hosts based on the `requested_destination` instruction
-        from the nova scheduler request spec. It supports filtering by host and
-        by aggregates.
----
-apiVersion: cortex.cloud/v1alpha1
-kind: Pipeline
-metadata:
-  name: nova-descheduler-kvm
-spec:
-  schedulingDomain: nova
-  description:
-    This pipeline runs steps that select virtual machines to deschedule from
-    compute hosts in order to optimize resource usage and performance.
-    This is the pipeline used for KVM hypervisors (qemu and cloud-hypervisor).
-  type: descheduler
-  {{- if $createDecisions }}
-  createDecisions: true
-  {{- end }}
-  steps:
-    - type: descheduler
-      impl: avoid_high_steal_pct
-      description: |
-        This step will deschedule VMs once they reach this CPU steal percentage over
-        the observed time span.
-      knowledges:
-        - name: kvm-libvirt-domain-cpu-steal-pct
-      opts:
-        maxStealPctOverObservedTimeSpan: 20.0
-      mandatory: false