cpu
×
Resource Quota
Overview
A ResourceQuota object enumerates hard resource usage limits per project. It limits the total number of a particular type of object that may be created in a project, and the total amount of compute resources that may be consumed by resources in that project.
Usage Limits
The following describes the set of limits that may be enforced by a ResourceQuota.
| Resource Name | Description |
|---|---|
Total requested cpu usage across all containers |
|
|
Total requested memory usage across all containers |
|
Total number of pods |
|
Total number of replication controllers |
|
Total number of resource quotas |
|
Total number of services |
|
Total number of secrets |
|
Total number of persistent volume claims |
Quota Enforcement
After a project quota is first created, the project restricts the ability to create any new resources that may violate a quota constraint until it has calculated updated usage statistics.
Once a quota is created and usage statistics are up-to-date, the project accepts the creation of new content. When you create or modify resources, your quota usage is incremented immediately upon the request to create or modify the resource. When you delete a resource, your quota use is decremented during the next full recalculation of quota statistics for the project. When you delete resources, a configurable amount of time determines how long it takes to reduce quota usage statistics to their current observed system value.
If project modifications exceed a quota usage limit, the server denies the action, and an appropriate error message is returned to the end-user explaining the quota constraint violated, and what their currently observed usage stats are in the system.
Compute Resources: Requests vs Limits
When allocating compute resources, each container may specify a request and a limit value for either CPU or memory. The quota tracking system will only cost the request value against the quota usage. If a resource is tracked by quota, and no request value is provided, the associated entity is rejected as part of admission into the cluster.
For an example, consider the following scenarios relative to tracking quota on CPU:
| Pod | Container | Requests[CPU] | Limits[CPU] | Result |
|---|---|---|---|---|
pod-x |
a |
100m |
500m |
The quota usage is incremented 100m |
pod-y |
b |
100m |
none |
The quota usage is incremented 100m |
pod-y |
c |
none |
500m |
The quota usage is incremented 500m since request will default to limit. |
pod-z |
d |
none |
none |
The pod is rejected since it does not enumerate a request. |
The rationale for charging quota usage for the requested amount of a resource versus the limit is the belief that a user should only be charged for what they are scheduled against in the cluster.
As a consequence, the user is able to spread its usage of a resource across multiple tiers of service. Let’s demonstrate this via an example with a 4 cpu quota.
The quota may be allocated as follows:
| Pod | Container | Requests[CPU] | Limits[CPU] | Quality of Service Tier | Quota Usage |
|---|---|---|---|---|---|
pod-x |
a |
1 |
4 |
Burstable |
1 |
pod-y |
b |
2 |
2 |
Guaranteed |
2 |
pod-z |
c |
1 |
3 |
Burstable |
1 |
The user is restricted from using BestEffort CPU containers, but at any point in time, the containers on the node may consume up to 9 CPU cores over a given time period if there is no contention on the node because it
has Burstable containers. If one wants to restrict the ratio between the request and limit, it is encouraged that the user define a LimitRange with MaxLimitRequestRatio to control burst out behavior. This would in effect, let an administrator keep the difference between request and limit more in line with tracked usage if desired.
Sample Resource Quota File
resource-quota.json
{
"apiVersion": "v1",
"kind": "ResourceQuota",
"metadata": {
"name": "quota" (1)
},
"spec": {
"hard": {
"memory": "1Gi", (2)
"cpu": "20", (3)
"pods": "10", (4)
"services": "5", (5)
"replicationcontrollers":"5", (6)
"resourcequotas":"1" (7)
}
}
}
| 1 | The name of this quota document |
| 2 | The total amount of memory requested across all containers may not exceed 1Gi. |
| 3 | The total number of cpu requested across all containers may not exceed 20 Kubernetes compute units. |
| 4 | The total number of pods in the project |
| 5 | The total number of services in the project |
| 6 | The total number of replication controllers in the project |
| 7 | The total number of resource quota documents in the project |
View a Quota
To view usage statistics related to any hard limits defined in your quota:
$ oc get quota NAME quota $ oc describe quota quota Name: quota Resource Used Hard -------- ---- ---- cpu 5 20 memory 500Mi 1Gi pods 5 10 replicationcontrollers 5 5 resourcequotas 1 1 services 3 5
Configuring the Quota Synchronization Period
When a set of resources are deleted, the synchronization timeframe of resources
is determined by the resource-quota-sync-period setting in the
/etc/origin/master/master-config.yaml file. Before your quota usage is
restored, you may encounter problems when attempting to reuse the resources.
Change the resource-quota-sync-period setting to have the set of resources
regenerate at the desired amount of time (in seconds) and for the resources to
be available again:
kubernetesMasterConfig:
apiLevels:
- v1beta3
- v1
apiServerArguments: null
controllerArguments:
resource-quota-sync-period:
- "10s"
Adjusting the regeneration time can be helpful for creating resources and determining resource usage when automation is used.
|
The |
Accounting for Quota in Deployment Configurations
If a quota has been defined for your project, see Deployment Resources for considerations on any deployment configurations.