Skip to content

docs: High Availability Setup documentation #2715

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 1 commit into
base: main
Choose a base branch
from
+90 −0
Open

docs: High Availability Setup documentation #2715

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
45 changes: 45 additions & 0 deletions README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -53,6 +53,7 @@ are deleted they are no longer visible on the `/metrics` endpoint.
* [Horizontal sharding](#horizontal-sharding)
* [Automated sharding](#automated-sharding)
* [Daemonset sharding for pod metrics](#daemonset-sharding-for-pod-metrics)
* [High Availability](#high-availability)
* [Setup](#setup)
* [Building the Docker container](#building-the-docker-container)
* [Usage](#usage)
Expand Down Expand Up @@ -304,6 +305,50 @@ spec:

Other metrics can be sharded via [Horizontal sharding](#horizontal-sharding).

### High Availability

Kube-state-metrics is a stateless service that reads from the Kubernetes API server. Be aware that multiple replicas increase the load on the Kubernetes API. Therefore, in most cases a single replica is also an option since most users scrape with a 30s interval. If you have the need for a higher scrape frequency or you have other constraints that require multiple replica, you can increase the availablity of kube-state-metrics in the following way:

For high availability, run multiple kube-state-metrics replicas to prevent a single point of failure. A common setup uses at least 2 replicas, pod anti-affinity rules to ensure they run on different nodes, and a PodDisruptionBudget (PDB) with `minAvailable: 1` to protect against voluntary disruptions.

```yaml
apiVersion: apps/v1
kind: Deployment
metadata:
name: kube-state-metrics
spec:
replicas: 2
template:
spec:
affinity:
podAntiAffinity:
preferredDuringSchedulingIgnoredDuringExecution:
- weight: 100
podAffinityTerm:
labelSelector:
matchLabels:
app: kube-state-metrics
topologyKey: kubernetes.io/hostname
containers:
- name: kube-state-metrics
image: registry.k8s.io/kube-state-metrics/kube-state-metrics:v2.10.0
ports:
- containerPort: 8080
name: http-metrics
---
apiVersion: policy/v1
kind: PodDisruptionBudget
metadata:
name: kube-state-metrics-pdb
spec:
minAvailable: 1
selector:
matchLabels:
app: kube-state-metrics
```

If metrics get scraped at the service level via a ServiceMonitor (Prometheus-Operator) or similar, metrics won't need to be deduplicated in a HA setup.

### Setup

Install this project to your `$GOPATH` using `go get`:
Expand Down
45 changes: 45 additions & 0 deletions README.md.tpl
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -53,6 +53,7 @@ are deleted they are no longer visible on the `/metrics` endpoint.
* [Horizontal sharding](#horizontal-sharding)
* [Automated sharding](#automated-sharding)
* [Daemonset sharding for pod metrics](#daemonset-sharding-for-pod-metrics)
* [High Availability](#high-availability)
* [Setup](#setup)
* [Building the Docker container](#building-the-docker-container)
* [Usage](#usage)
Expand Down Expand Up @@ -305,6 +306,50 @@ spec:

Other metrics can be sharded via [Horizontal sharding](#horizontal-sharding).

### High Availability

Kube-state-metrics is a stateless service that reads from the Kubernetes API server. Be aware that multiple replicas increase the load on the Kubernetes API. Therefore, in most cases a single replica is also an option since most users scrape with a 30s interval. If you have the need for a higher scrape frequency or you have other constraints that require multiple replica, you can increase the availablity of kube-state-metrics in the following way:

For high availability, run multiple kube-state-metrics replicas to prevent a single point of failure. A common setup uses at least 2 replicas, pod anti-affinity rules to ensure they run on different nodes, and a PodDisruptionBudget (PDB) with `minAvailable: 1` to protect against voluntary disruptions.

```yaml
apiVersion: apps/v1
kind: Deployment
metadata:
name: kube-state-metrics
spec:
replicas: 2
template:
spec:
affinity:
podAntiAffinity:
preferredDuringSchedulingIgnoredDuringExecution:
- weight: 100
podAffinityTerm:
labelSelector:
matchLabels:
app: kube-state-metrics
topologyKey: kubernetes.io/hostname
containers:
- name: kube-state-metrics
image: registry.k8s.io/kube-state-metrics/kube-state-metrics:v2.10.0
ports:
- containerPort: 8080
name: http-metrics
---
apiVersion: policy/v1
kind: PodDisruptionBudget
metadata:
name: kube-state-metrics-pdb
spec:
minAvailable: 1
selector:
matchLabels:
app: kube-state-metrics
```

If metrics get scraped at the service level via a ServiceMonitor (Prometheus-Operator) or similar, metrics won't need to be deduplicated in a HA setup.

### Setup

Install this project to your `$GOPATH` using `go get`:
Expand Down