Chapter 8. Using CPU Manager and Topology Manager
CPU Manager manages groups of CPUs and constrains workloads to specific CPUs.
CPU Manager is useful for workloads that have some of these attributes:
- Require as much CPU time as possible.
- Are sensitive to processor cache misses.
- Are low-latency network applications.
- Coordinate with other processes and benefit from sharing a single processor cache.
Topology Manager collects hints from the CPU Manager, Device Manager, and other Hint Providers to align pod resources, such as CPU, SR-IOV VFs, and other device resources, for all Quality of Service (QoS) classes on the same non-uniform memory access (NUMA) node.
Topology Manager uses topology information from the collected hints to decide if a pod can be accepted or rejected on a node, based on the configured Topology Manager policy and pod resources requested.
Topology Manager is useful for workloads that use hardware accelerators to support latency-critical execution and high throughput parallel computation.
To use Topology Manager you must configure CPU Manager with the static
policy.
8.1. Setting up CPU Manager
To configure CPU manager, create a KubeletConfig custom resource (CR) and apply it to the desired set of nodes.
Procedure
Label a node by running the following command:
# oc label node perf-node.example.com cpumanager=true
To enable CPU Manager for all compute nodes, edit the CR by running the following command:
# oc edit machineconfigpool worker
Add the
custom-kubelet: cpumanager-enabled
label tometadata.labels
section.metadata: creationTimestamp: 2020-xx-xxx generation: 3 labels: custom-kubelet: cpumanager-enabled
Create a
KubeletConfig
,cpumanager-kubeletconfig.yaml
, custom resource (CR). Refer to the label created in the previous step to have the correct nodes updated with the new kubelet config. See themachineConfigPoolSelector
section:apiVersion: machineconfiguration.openshift.io/v1 kind: KubeletConfig metadata: name: cpumanager-enabled spec: machineConfigPoolSelector: matchLabels: custom-kubelet: cpumanager-enabled kubeletConfig: cpuManagerPolicy: static 1 cpuManagerReconcilePeriod: 5s 2
- 1
- Specify a policy:
-
none
. This policy explicitly enables the existing default CPU affinity scheme, providing no affinity beyond what the scheduler does automatically. This is the default policy. -
static
. This policy allows containers in guaranteed pods with integer CPU requests. It also limits access to exclusive CPUs on the node. Ifstatic
, you must use a lowercases
.
-
- 2
- Optional. Specify the CPU Manager reconcile frequency. The default is
5s
.
Create the dynamic kubelet config by running the following command:
# oc create -f cpumanager-kubeletconfig.yaml
This adds the CPU Manager feature to the kubelet config and, if needed, the Machine Config Operator (MCO) reboots the node. To enable CPU Manager, a reboot is not needed.
Check for the merged kubelet config by running the following command:
# oc get machineconfig 99-worker-XXXXXX-XXXXX-XXXX-XXXXX-kubelet -o json | grep ownerReference -A7
Example output
"ownerReferences": [ { "apiVersion": "machineconfiguration.openshift.io/v1", "kind": "KubeletConfig", "name": "cpumanager-enabled", "uid": "7ed5616d-6b72-11e9-aae1-021e1ce18878" } ]
Check the compute node for the updated
kubelet.conf
file by running the following command:# oc debug node/perf-node.example.com sh-4.2# cat /host/etc/kubernetes/kubelet.conf | grep cpuManager
Example output
cpuManagerPolicy: static 1 cpuManagerReconcilePeriod: 5s 2
Create a project by running the following command:
$ oc new-project <project_name>
Create a pod that requests a core or multiple cores. Both limits and requests must have their CPU value set to a whole integer. That is the number of cores that will be dedicated to this pod:
# cat cpumanager-pod.yaml
Example output
apiVersion: v1 kind: Pod metadata: generateName: cpumanager- spec: securityContext: runAsNonRoot: true seccompProfile: type: RuntimeDefault containers: - name: cpumanager image: gcr.io/google_containers/pause:3.2 resources: requests: cpu: 1 memory: "1G" limits: cpu: 1 memory: "1G" securityContext: allowPrivilegeEscalation: false capabilities: drop: [ALL] nodeSelector: cpumanager: "true"
Create the pod:
# oc create -f cpumanager-pod.yaml
Verification
Verify that the pod is scheduled to the node that you labeled by running the following command:
# oc describe pod cpumanager
Example output
Name: cpumanager-6cqz7 Namespace: default Priority: 0 PriorityClassName: <none> Node: perf-node.example.com/xxx.xx.xx.xxx ... Limits: cpu: 1 memory: 1G Requests: cpu: 1 memory: 1G ... QoS Class: Guaranteed Node-Selectors: cpumanager=true
Verify that a CPU has been exclusively assigned to the pod by running the following command:
# oc describe node --selector='cpumanager=true' | grep -i cpumanager- -B2
Example output
NAMESPACE NAME CPU Requests CPU Limits Memory Requests Memory Limits Age cpuman cpumanager-mlrrz 1 (28%) 1 (28%) 1G (13%) 1G (13%) 27m
Verify that the
cgroups
are set up correctly. Get the process ID (PID) of thepause
process by running the following commands:# oc debug node/perf-node.example.com
sh-4.2# systemctl status | grep -B5 pause
NoteIf the output returns multiple pause process entries, you must identify the correct pause process.
Example output
# ├─init.scope │ └─1 /usr/lib/systemd/systemd --switched-root --system --deserialize 17 └─kubepods.slice ├─kubepods-pod69c01f8e_6b74_11e9_ac0f_0a2b62178a22.slice │ ├─crio-b5437308f1a574c542bdf08563b865c0345c8f8c0b0a655612c.scope │ └─32706 /pause
Verify that pods of quality of service (QoS) tier
Guaranteed
are placed within thekubepods.slice
subdirectory by running the following commands:# cd /sys/fs/cgroup/kubepods.slice/kubepods-pod69c01f8e_6b74_11e9_ac0f_0a2b62178a22.slice/crio-b5437308f1ad1a7db0574c542bdf08563b865c0345c86e9585f8c0b0a655612c.scope
# for i in `ls cpuset.cpus cgroup.procs` ; do echo -n "$i "; cat $i ; done
NotePods of other QoS tiers end up in child
cgroups
of the parentkubepods
.Example output
cpuset.cpus 1 tasks 32706
Check the allowed CPU list for the task by running the following command:
# grep ^Cpus_allowed_list /proc/32706/status
Example output
Cpus_allowed_list: 1
Verify that another pod on the system cannot run on the core allocated for the
Guaranteed
pod. For example, to verify the pod in thebesteffort
QoS tier, run the following commands:# cat /sys/fs/cgroup/kubepods.slice/kubepods-besteffort.slice/kubepods-besteffort-podc494a073_6b77_11e9_98c0_06bba5c387ea.slice/crio-c56982f57b75a2420947f0afc6cafe7534c5734efc34157525fa9abbf99e3849.scope/cpuset.cpus
# oc describe node perf-node.example.com
Example output
... Capacity: attachable-volumes-aws-ebs: 39 cpu: 2 ephemeral-storage: 124768236Ki hugepages-1Gi: 0 hugepages-2Mi: 0 memory: 8162900Ki pods: 250 Allocatable: attachable-volumes-aws-ebs: 39 cpu: 1500m ephemeral-storage: 124768236Ki hugepages-1Gi: 0 hugepages-2Mi: 0 memory: 7548500Ki pods: 250 ------- ---- ------------ ---------- --------------- ------------- --- default cpumanager-6cqz7 1 (66%) 1 (66%) 1G (12%) 1G (12%) 29m Allocated resources: (Total limits may be over 100 percent, i.e., overcommitted.) Resource Requests Limits -------- -------- ------ cpu 1440m (96%) 1 (66%)
This VM has two CPU cores. The
system-reserved
setting reserves 500 millicores, meaning that half of one core is subtracted from the total capacity of the node to arrive at theNode Allocatable
amount. You can see thatAllocatable CPU
is 1500 millicores. This means you can run one of the CPU Manager pods since each will take one whole core. A whole core is equivalent to 1000 millicores. If you try to schedule a second pod, the system will accept the pod, but it will never be scheduled:NAME READY STATUS RESTARTS AGE cpumanager-6cqz7 1/1 Running 0 33m cpumanager-7qc2t 0/1 Pending 0 11s
8.2. Topology Manager policies
Topology Manager aligns Pod
resources of all Quality of Service (QoS) classes by collecting topology hints from Hint Providers, such as CPU Manager and Device Manager, and using the collected hints to align the Pod
resources.
Topology Manager supports four allocation policies, which you assign in the KubeletConfig
custom resource (CR) named cpumanager-enabled
:
none
policy- This is the default policy and does not perform any topology alignment.
best-effort
policy-
For each container in a pod with the
best-effort
topology management policy, kubelet calls each Hint Provider to discover their resource availability. Using this information, the Topology Manager stores the preferred NUMA Node affinity for that container. If the affinity is not preferred, Topology Manager stores this and admits the pod to the node. restricted
policy-
For each container in a pod with the
restricted
topology management policy, kubelet calls each Hint Provider to discover their resource availability. Using this information, the Topology Manager stores the preferred NUMA Node affinity for that container. If the affinity is not preferred, Topology Manager rejects this pod from the node, resulting in a pod in aTerminated
state with a pod admission failure. single-numa-node
policy-
For each container in a pod with the
single-numa-node
topology management policy, kubelet calls each Hint Provider to discover their resource availability. Using this information, the Topology Manager determines if a single NUMA Node affinity is possible. If it is, the pod is admitted to the node. If a single NUMA Node affinity is not possible, the Topology Manager rejects the pod from the node. This results in a pod in a Terminated state with a pod admission failure.
8.3. Setting up Topology Manager
To use Topology Manager, you must configure an allocation policy in the KubeletConfig
custom resource (CR) named cpumanager-enabled
. This file might exist if you have set up CPU Manager. If the file does not exist, you can create the file.
Prerequisites
-
Configure the CPU Manager policy to be
static
.
Procedure
To activate Topology Manager:
Configure the Topology Manager allocation policy in the custom resource.
$ oc edit KubeletConfig cpumanager-enabled
apiVersion: machineconfiguration.openshift.io/v1 kind: KubeletConfig metadata: name: cpumanager-enabled spec: machineConfigPoolSelector: matchLabels: custom-kubelet: cpumanager-enabled kubeletConfig: cpuManagerPolicy: static 1 cpuManagerReconcilePeriod: 5s topologyManagerPolicy: single-numa-node 2
8.4. Pod interactions with Topology Manager policies
The example Pod
specs below help illustrate pod interactions with Topology Manager.
The following pod runs in the BestEffort
QoS class because no resource requests or limits are specified.
spec: containers: - name: nginx image: nginx
The next pod runs in the Burstable
QoS class because requests are less than limits.
spec: containers: - name: nginx image: nginx resources: limits: memory: "200Mi" requests: memory: "100Mi"
If the selected policy is anything other than none
, Topology Manager would not consider either of these Pod
specifications.
The last example pod below runs in the Guaranteed QoS class because requests are equal to limits.
spec: containers: - name: nginx image: nginx resources: limits: memory: "200Mi" cpu: "2" example.com/device: "1" requests: memory: "200Mi" cpu: "2" example.com/device: "1"
Topology Manager would consider this pod. The Topology Manager would consult the hint providers, which are CPU Manager and Device Manager, to get topology hints for the pod.
Topology Manager will use this information to store the best topology for this container. In the case of this pod, CPU Manager and Device Manager will use this stored information at the resource allocation stage.