Chapter 4. Limits and scalability
This document details the tested cluster maximums for Red Hat OpenShift Service on AWS (ROSA) clusters, along with information about the test environment and configuration used to test the maximums. Information about control plane and infrastructure node sizing and scaling is also provided.
4.1. Cluster maximums
Consider the following tested object maximums when you plan a Red Hat OpenShift Service on AWS (ROSA) cluster installation. The table specifies the maximum limits for each tested type in a (ROSA) cluster.
These guidelines are based on a cluster of 102 compute (also known as worker) nodes in a multiple availability zone configuration. For smaller clusters, the maximums are lower.
The OpenShift Container Platform version used in all of the tests is OCP 4.8.0.
Table 4.1. Tested cluster maximums
| Maximum type | 4.8 tested maximum |
|---|---|
| Number of nodes | 102 |
| Number of pods [1] | 20,400 |
| Number of pods per node | 250 |
| Number of pods per core | There is no default value |
| Number of namespaces [2] | 3,400 |
| Number of pods per namespace [3] | 20,400 |
| Number of services [4] | 10,000 |
| Number of services per namespace | 10,000 |
| Number of back ends per service | 10,000 |
| Number of deployments per namespace [3] | 1,000 |
- The pod count displayed here is the number of test pods. The actual number of pods depends on the application’s memory, CPU, and storage requirements.
- When there are a large number of active projects, etcd can suffer from poor performance if the keyspace grows excessively large and exceeds the space quota. Periodic maintenance of etcd, including defragmentation, is highly recommended to make etcd storage available.
- There are a number of control loops in the system that must iterate over all objects in a given namespace as a reaction to some changes in state. Having a large number of objects of a type, in a single namespace, can make those loops expensive and slow down processing the state changes. The limit assumes that the system has enough CPU, memory, and disk to satisfy the application requirements.
- Each service port and each service back end has a corresponding entry in iptables. The number of back ends of a given service impacts the size of the endpoints objects, which then impacts the size of data that is sent throughout the system.
In OpenShift Container Platform 4.8, half of a CPU core (500 millicore) is reserved by the system compared to previous versions of OpenShift Container Platform.
4.2. OpenShift Container Platform testing environment and configuration
The following table lists the OpenShift Container Platform environment and configuration on which the cluster maximums are tested for the AWS cloud platform.
| Node | Type | vCPU | RAM(GiB) | Disk type | Disk size(GiB)/IOPS | Count | Region |
|---|---|---|---|---|---|---|---|
| Control plane/etcd [1] | m5.4xlarge | 16 | 64 | gp3 | 350 / 1,000 | 3 | us-west-2 |
| Infrastructure nodes [2] | r5.2xlarge | 8 | 64 | gp3 | 300 / 900 | 3 | us-west-2 |
| Workload [3] | m5.2xlarge | 8 | 32 | gp3 | 350 / 900 | 3 | us-west-2 |
| Compute nodes | m5.2xlarge | 8 | 32 | gp3 | 350 / 900 | 102 | us-west-2 |
- io1 disks are used for control plane/etcd nodes in all versions prior to 4.10.
- Infrastructure nodes are used to host monitoring components because Prometheus can claim a large amount of memory, depending on usage patterns.
- Workload nodes are dedicated to run performance and scalability workload generators.
Larger cluster sizes and higher object counts might be reachable. However, the sizing of the infrastructure nodes limits the amount of memory that is available to Prometheus. When creating, modifying, or deleting objects, Prometheus stores the metrics in its memory for roughly 3 hours prior to persisting the metrics on disk. If the rate of creation, modification, or deletion of objects is too high, Prometheus can become overwhelmed and fail due to the lack of memory resources.
4.3. Control plane and infrastructure node sizing and scaling
When you install a Red Hat OpenShift Service on AWS (ROSA) cluster, the sizing of the control plane and infrastructure nodes are automatically determined by the compute node count.
If you change the number of compute nodes in your cluster after installation, the Red Hat Site Reliability Engineering (SRE) team scales the control plane and infrastructure nodes as required to maintain cluster stability.
4.3.1. Node sizing during installation
During the installation process, the sizing of the control plane and infrastructure nodes are dynamically calculated. The sizing calculation is based on the number of compute nodes in a cluster.
The following table lists the control plane and infrastructure node sizing that is applied during installation.
| Number of compute nodes | Control plane size | Infrastructure node size |
|---|---|---|
| 1 to 25 | m5.2xlarge | r5.xlarge |
| 26 to 100 | m5.4xlarge | r5.2xlarge |
| 101 to 180 | m5.8xlarge | r5.4xlarge |
The maximum number of compute nodes on ROSA is 180.
4.3.2. Node scaling after installation
If you change the number of compute nodes after installation, the control plane and infrastructure nodes are scaled by the Red Hat Site Reliability Engineering (SRE) team as required. The nodes are scaled to maintain platform stability.
Postinstallation scaling requirements for control plane and infrastructure nodes are assessed on a case-by-case basis. Node resource consumption and received alerts are taken into consideration.
Rules for control plane node resizing alerts
The resizing alert is triggered for the control plane nodes in a cluster when the following occurs:
Control plane nodes sustain over 66% utilization on average in a classic ROSA cluster.
NoteThe maximum number of compute nodes on ROSA is 180.
Rules for infrastructure node resizing alerts
Resizing alerts are triggered for the infrastructure nodes in a cluster when it has high-sustained CPU or memory utilization. This high-sustained utilization status is:
- Infrastructure nodes sustain over 50% utilization on average in a classic ROSA cluster with a single availability zone using 2 infrastructure nodes.
Infrastructure nodes sustain over 66% utilization on average in a classic ROSA cluster with multiple availability zones using 3 infrastructure nodes.
NoteThe maximum number of compute nodes on ROSA is 180.
The resizing alerts only appear after sustained periods of high utilization. Short usage spikes, such as a node temporarily going down causing the other node to scale up, do not trigger these alerts.
The SRE team might scale the control plane and infrastructure nodes for additional reasons, for example to manage an increase in resource consumption on the nodes.
When scaling is applied, the customer is notified through a service log entry. For more information about the service log, see Accessing the service logs for ROSA clusters.
4.3.3. Sizing considerations for larger clusters
For larger clusters, infrastructure node sizing can become a significant impacting factor to scalability. There are many factors that influence the stated thresholds, including the etcd version or storage data format.
Exceeding these limits does not necessarily mean that the cluster will fail. In most cases, exceeding these numbers results in lower overall performance.
