Skip to content

Commit

Permalink
Introduce InstanceTypes
Browse files Browse the repository at this point in the history
Signed-off-by: Andrei Kvapil <[email protected]>
  • Loading branch information
kvaps committed Oct 10, 2024
1 parent 16d3979 commit 7f623c5
Show file tree
Hide file tree
Showing 18 changed files with 4,216 additions and 89 deletions.
4 changes: 4 additions & 0 deletions packages/apps/kubernetes/Makefile
Original file line number Diff line number Diff line change
Expand Up @@ -65,3 +65,7 @@ image-cluster-autoscaler:
echo "$(REGISTRY)/cluster-autoscaler:$(call settag,$(KUBERNETES_PKG_TAG))@$$(yq e '."containerimage.digest"' images/cluster-autoscaler.json -o json -r)" \
> images/cluster-autoscaler.tag
rm -f images/cluster-autoscaler.json

prefs:
INSTANCE_TYPES=$(yq e '.metadata.name' -o=json -r ../../system/kubevirt-common-instancetypes/templates/instancetypes.yaml)
INSTANCE_PREFS=$(yq e '.metadata.name' -o=json -r ../../system/kubevirt-common-instancetypes/templates/preferences.yaml)
263 changes: 263 additions & 0 deletions packages/apps/kubernetes/README.md
Original file line number Diff line number Diff line change
Expand Up @@ -50,3 +50,266 @@ kubectl get secret -n <namespace> kubernetes-<clusterName>-admin-kubeconfig -o g
| `addons.fluxcd.enabled` | Enables Flux CD | `false` |
| `addons.fluxcd.valuesOverride` | Custom values to override | `{}` |

## U Series

The U Series is quite neutral and provides resources for
general purpose applications.

*U* is the abbreviation for "Universal", hinting at the universal
attitude towards workloads.

VMs of instance types will share physical CPU cores on a
time-slice basis with other VMs.

### U Series Characteristics

Specific characteristics of this series are:
- *Burstable CPU performance* - The workload has a baseline compute
performance but is permitted to burst beyond this baseline, if
excess compute resources are available.
- *vCPU-To-Memory Ratio (1:4)* - A vCPU-to-Memory ratio of 1:4, for less
noise per node.

## O Series

The O Series is based on the U Series, with the only difference
being that memory is overcommitted.

*O* is the abbreviation for "Overcommitted".

### UO Series Characteristics

Specific characteristics of this series are:
- *Burstable CPU performance* - The workload has a baseline compute
performance but is permitted to burst beyond this baseline, if
excess compute resources are available.
- *Overcommitted Memory* - Memory is over-committed in order to achieve
a higher workload density.
- *vCPU-To-Memory Ratio (1:4)* - A vCPU-to-Memory ratio of 1:4, for less
noise per node.

## CX Series

The CX Series provides exclusive compute resources for compute
intensive applications.

*CX* is the abbreviation of "Compute Exclusive".

The exclusive resources are given to the compute threads of the
VM. In order to ensure this, some additional cores (depending
on the number of disks and NICs) will be requested to offload
the IO threading from cores dedicated to the workload.
In addition, in this series, the NUMA topology of the used
cores is provided to the VM.

### CX Series Characteristics

Specific characteristics of this series are:
- *Hugepages* - Hugepages are used in order to improve memory
performance.
- *Dedicated CPU* - Physical cores are exclusively assigned to every
vCPU in order to provide fixed and high compute guarantees to the
workload.
- *Isolated emulator threads* - Hypervisor emulator threads are isolated
from the vCPUs in order to reduce emaulation related impact on the
workload.
- *vNUMA* - Physical NUMA topology is reflected in the guest in order to
optimize guest sided cache utilization.
- *vCPU-To-Memory Ratio (1:2)* - A vCPU-to-Memory ratio of 1:2.

## GN Series

The GN Series provides instances types intended for VMs with
NVIDIA GPU resources attached.

*GN* is the abbreviation of "GPU NVIDIA".

This series is intended to be used with VMs consuming GPUs
provided by the
[NVIDIA GPU Operator](https://github.com/NVIDIA/gpu-operator)
which can be installed on Kubernetes and also is made available
on OpenShift via OperatorHub.

### GN Series Characteristics

Specific characteristics of this series are:
- *Has GPUs* - Has GPUs predefined.
- *Burstable CPU performance* - The workload has a baseline compute
performance but is permitted to burst beyond this baseline, if
excess compute resources are available.
- *vCPU-To-Memory Ratio (1:4)* - A vCPU-to-Memory ratio of 1:4, for less
noise per node.

## M Series

The M Series provides resources for memory intensive
applications.

*M* is the abbreviation of "Memory".

### M Series Characteristics

Specific characteristics of this series are:
- *Hugepages* - Hugepages are used in order to improve memory
performance.
- *Burstable CPU performance* - The workload has a baseline compute
performance but is permitted to burst beyond this baseline, if
excess compute resources are available.
- *vCPU-To-Memory Ratio (1:8)* - A vCPU-to-Memory ratio of 1:8, for much
less noise per node.

## N Series

The N Series provides resources for network intensive DPDK
applications, like VNFs.

*N* is the abbreviation for "Network".

This series of instancetypes requires nodes capable
of running DPDK workloads and being marked with the respective
node-role.kubevirt.io/worker-dpdk label as such.

### N Series Characteristics

Specific characteristics of this series are:
- *Hugepages* - Hugepages are used in order to improve memory
performance.
- *Dedicated CPU* - Physical cores are exclusively assigned to every
vCPU in order to provide fixed and high compute guarantees to the
workload.
- *Isolated emulator threads* - Hypervisor emulator threads are isolated
from the vCPUs in order to reduce emaulation related impact on the
workload.
- *vCPU-To-Memory Ratio (1:2)* - A vCPU-to-Memory ratio of 1:2.

## RT Series

The RT Series provides resources for realtime applications, like Oslat.

*RT* is the abbreviation for "realtime".

This series of instance types requires nodes capable of running
realtime applications.

### RT Series Characteristics

Specific characteristics of this series are:
- *Hugepages* - Hugepages are used in order to improve memory
performance.
- *Dedicated CPU* - Physical cores are exclusively assigned to every
vCPU in order to provide fixed and high compute guarantees to the
workload.
- *Isolated emulator threads* - Hypervisor emulator threads are isolated
from the vCPUs in order to reduce emaulation related impact on the
workload.
- *vCPU-To-Memory Ratio (1:4)* - A vCPU-to-Memory ratio of 1:4 starting from
the medium size.

## Development

To get started with customizing or creating your own instancetypes and preferences
see [DEVELOPMENT.md](./DEVELOPMENT.md).

## Resources

The following instancetype resources (cluster-wide and namespaced) are
provided by this project:

Name | vCPUs | Memory
-----|-------|-------
cx1.2xlarge | 8 | 16Gi
cx1.4xlarge | 16 | 32Gi
cx1.8xlarge | 32 | 64Gi
cx1.large | 2 | 4Gi
cx1.medium | 1 | 2Gi
cx1.xlarge | 4 | 8Gi
gn1.2xlarge | 8 | 32Gi
gn1.4xlarge | 16 | 64Gi
gn1.8xlarge | 32 | 128Gi
gn1.xlarge | 4 | 16Gi
m1.2xlarge | 8 | 64Gi
m1.4xlarge | 16 | 128Gi
m1.8xlarge | 32 | 256Gi
m1.large | 2 | 16Gi
m1.xlarge | 4 | 32Gi
n1.2xlarge | 16 | 32Gi
n1.4xlarge | 32 | 64Gi
n1.8xlarge | 64 | 128Gi
n1.large | 4 | 8Gi
n1.medium | 4 | 4Gi
n1.xlarge | 8 | 16Gi
o1.2xlarge | 8 | 32Gi
o1.4xlarge | 16 | 64Gi
o1.8xlarge | 32 | 128Gi
o1.large | 2 | 8Gi
o1.medium | 1 | 4Gi
o1.micro | 1 | 1Gi
o1.nano | 1 | 512Mi
o1.small | 1 | 2Gi
o1.xlarge | 4 | 16Gi
rt1.2xlarge | 8 | 32Gi
rt1.4xlarge | 16 | 64Gi
rt1.8xlarge | 32 | 128Gi
rt1.large | 2 | 8Gi
rt1.medium | 1 | 4Gi
rt1.micro | 1 | 1Gi
rt1.small | 1 | 2Gi
rt1.xlarge | 4 | 16Gi
u1.2xlarge | 8 | 32Gi
u1.2xmedium | 2 | 4Gi
u1.4xlarge | 16 | 64Gi
u1.8xlarge | 32 | 128Gi
u1.large | 2 | 8Gi
u1.medium | 1 | 4Gi
u1.micro | 1 | 1Gi
u1.nano | 1 | 512Mi
u1.small | 1 | 2Gi
u1.xlarge | 4 | 16Gi

The following preference resources (cluster-wide and namespaced) are
provided by this project:

Name | Guest OS
-----|---------
alpine | Alpine
centos.7 | CentOS 7
centos.7.desktop | CentOS 7
centos.stream10 | CentOS Stream 10
centos.stream10.desktop | CentOS Stream 10
centos.stream8 | CentOS Stream 8
centos.stream8.desktop | CentOS Stream 8
centos.stream8.dpdk | CentOS Stream 8
centos.stream9 | CentOS Stream 9
centos.stream9.desktop | CentOS Stream 9
centos.stream9.dpdk | CentOS Stream 9
cirros | Cirros
fedora | Fedora (amd64)
fedora.arm64 | Fedora (arm64)
opensuse.leap | OpenSUSE Leap
opensuse.tumbleweed | OpenSUSE Tumbleweed
rhel.10 | Red Hat Enterprise Linux 10 Beta (amd64)
rhel.10.arm64 | Red Hat Enterprise Linux 10 Beta (arm64)
rhel.7 | Red Hat Enterprise Linux 7
rhel.7.desktop | Red Hat Enterprise Linux 7
rhel.8 | Red Hat Enterprise Linux 8
rhel.8.desktop | Red Hat Enterprise Linux 8
rhel.8.dpdk | Red Hat Enterprise Linux 8
rhel.9 | Red Hat Enterprise Linux 9 (amd64)
rhel.9.arm64 | Red Hat Enterprise Linux 9 (arm64)
rhel.9.desktop | Red Hat Enterprise Linux 9 Desktop (amd64)
rhel.9.dpdk | Red Hat Enterprise Linux 9 DPDK (amd64)
rhel.9.realtime | Red Hat Enterprise Linux 9 Realtime (amd64)
sles | SUSE Linux Enterprise Server
ubuntu | Ubuntu
windows.10 | Microsoft Windows 10
windows.10.virtio | Microsoft Windows 10 (virtio)
windows.11 | Microsoft Windows 11
windows.11.virtio | Microsoft Windows 11 (virtio)
windows.2k16 | Microsoft Windows Server 2016
windows.2k16.virtio | Microsoft Windows Server 2016 (virtio)
windows.2k19 | Microsoft Windows Server 2019
windows.2k19.virtio | Microsoft Windows Server 2019 (virtio)
windows.2k22 | Microsoft Windows Server 2022
windows.2k22.virtio | Microsoft Windows Server 2022 (virtio)
windows.2k25 | Microsoft Windows Server 2025
windows.2k25.virtio | Microsoft Windows Server 2025 (virtio)
25 changes: 25 additions & 0 deletions packages/apps/kubernetes/templates/cluster.yaml
Original file line number Diff line number Diff line change
Expand Up @@ -15,6 +15,11 @@ spec:
node-role.kubernetes.io/{{ . }}: ""
{{- end }}
spec:
{{- with .group.instanceType }}
instancetype:
kind: VirtualMachineClusterInstancetype
name: {{ . }}
{{- end }}
runStrategy: Always
template:
metadata:
Expand All @@ -26,10 +31,12 @@ spec:
{{- end }}
spec:
domain:
{{- if and .group.resources .group.resources.cpu }}
cpu:
threads: 1
cores: {{ .group.resources.cpu }}
sockets: 1
{{- end }}
devices:
disks:
- name: system
Expand All @@ -43,8 +50,10 @@ spec:
interfaces:
- name: default
bridge: {}
{{- if and .group.resources .group.resources.memory }}
memory:
guest: {{ .group.resources.memory }}
{{- end }}
evictionStrategy: External
volumes:
- name: system
Expand Down Expand Up @@ -176,6 +185,14 @@ spec:
template:
{{- $kubevirtmachinetemplate | nindent 4 }}
---
{{- $instanceType := dict }}
{{- if $group.instanceType }}
{{- $instanceType = (lookup "instancetype.kubevirt.io/v1beta1" "VirtualMachineClusterInstancetype" "" $group.instanceType) }}
{{- if not $instanceType }}
{{- fail (printf "Specified instancetype not exists in cluster: %s" $group.instanceType) }}
{{- end }}
{{- end }}

apiVersion: cluster.x-k8s.io/v1beta1
kind: MachineDeployment
metadata:
Expand All @@ -184,8 +201,16 @@ metadata:
annotations:
cluster.x-k8s.io/cluster-api-autoscaler-node-group-min-size: "{{ $group.minReplicas }}"
cluster.x-k8s.io/cluster-api-autoscaler-node-group-max-size: "{{ $group.maxReplicas }}"
{{- if and $group.resources $group.resources.memory }}
capacity.cluster-autoscaler.kubernetes.io/memory: "{{ $group.resources.memory }}"
{{- else }}
capacity.cluster-autoscaler.kubernetes.io/memory: "{{ $instanceType.spec.memory.guest }}"
{{- end }}
{{- if and $group.resources $group.resources.cpu }}
capacity.cluster-autoscaler.kubernetes.io/cpu: "{{ $group.resources.cpu }}"
{{- else }}
capacity.cluster-autoscaler.kubernetes.io/cpu: "{{ $instanceType.spec.cpu.guest }}"
{{- end }}
spec:
clusterName: {{ $.Release.Name }}
template:
Expand Down
8 changes: 5 additions & 3 deletions packages/apps/kubernetes/values.yaml
Original file line number Diff line number Diff line change
Expand Up @@ -15,13 +15,15 @@ nodeGroups:
md0:
minReplicas: 0
maxReplicas: 10
resources:
cpu: 2
memory: 1024Mi
instanceType: "u1.medium"
ephemeralStorage: 20Gi
roles:
- ingress-nginx

resources:
cpu: ""
memory: ""

## @section Cluster Addons
##
addons:
Expand Down
2 changes: 1 addition & 1 deletion packages/apps/virtual-machine/Chart.yaml
Original file line number Diff line number Diff line change
Expand Up @@ -17,7 +17,7 @@ type: application
# This is the chart version. This version number should be incremented each time you make changes
# to the chart and its templates, including the app version.
# Versions are expected to follow Semantic Versioning (https://semver.org/)
version: 0.4.0
version: 0.5.0

# This is the version number of the application being deployed. This version number should be
# incremented each time you make changes to the application. Versions are not expected to
Expand Down
13 changes: 6 additions & 7 deletions packages/apps/virtual-machine/Makefile
Original file line number Diff line number Diff line change
@@ -1,10 +1,9 @@
include ../../../scripts/package.mk

generate:
readme-generator -v values.yaml -s values.schema.json.tmp -r README.md
cat values.schema.json.tmp | \
jq '.properties.image.enum = ["ubuntu", "cirros", "alpine", "fedora", "talos"]' | \
jq '.properties.resources.properties.memory["x-display"] = "slider"' | \
jq '.properties.externalPorts.items.type = "integer"' \
> values.schema.json
rm -f values.schema.json.tmp
readme-generator -v values.yaml -s values.schema.json -r README.md
INSTANCE_TYPES=$$(yq e '.metadata.name' -o=json -r ../../system/kubevirt-common-instancetypes/templates/instancetypes.yaml | yq 'split(" ")' -o json) \
&& yq -i -o json ".properties.instanceType.optional=true | .properties.instanceType.enum = $${INSTANCE_TYPES}" values.schema.json
PREFERENCES=$$(yq e '.metadata.name' -o=json -r ../../system/kubevirt-common-instancetypes/templates/preferences.yaml | yq 'split(" ")' -o json) \
&& yq -i -o json ".properties.instanceProfile.optional=true | .properties.instanceProfile.enum = $${PREFERENCES}" values.schema.json
yq -i -o json '.properties.externalPorts.items.type = "integer"' values.schema.json
Loading

0 comments on commit 7f623c5

Please sign in to comment.