Skip to content

Commit 81101f1

Browse files
shubhusionshubhusion
authored
Created Procedure for Setting Up Autoscaling with KEDA (#1446)
1 parent 03b09f9 commit 81101f1

File tree

3 files changed

+360
-0
lines changed

3 files changed

+360
-0
lines changed

‎content/docs/2.14/setupscaler.md

+120
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,120 @@
1+
+++
2+
title = "Setup Autoscaling with KEDA"
3+
weight = 500
4+
description = "Procedure to Setup a Scaler in KEDA"
5+
+++
6+
7+
## Prerequisites
8+
9+
1. **Kubernetes Cluster**:
10+
- Ensure you have a running Kubernetes cluster set up and accessible.
11+
- If you don't have a cluster yet, follow the [official Kubernetes documentation](https://kubernetes.io/docs/setup/) to create a new cluster suitable for your environment (local machine, cloud provider, etc.).
12+
13+
2. **KEDA Installation**:
14+
- KEDA needs to be installed on your Kubernetes cluster before you can use it.
15+
- Follow the [KEDA installation guide](https://keda.sh/docs/2.14/deploy/) carefully, including any prerequisites specific to your Kubernetes setup.
16+
- The installation guide provides instructions for different installation methods (e.g., YAML, Helm charts, etc.). Choose the method that suits your needs.
17+
18+
3. **kubectl**:
19+
- The `kubectl` command-line tool is required to interact with your Kubernetes cluster.
20+
- Follow the [official kubectl installation guide](https://kubernetes.io/docs/tasks/tools/#kubectl) to install `kubectl` on your operating system.
21+
- Once installed, configure `kubectl` to communicate with your Kubernetes cluster by following the cluster-specific instructions provided by your Kubernetes setup.
22+
23+
## Step 1: Identify the Scaler You Need
24+
25+
KEDA supports various scalers that correspond to different event sources or triggers. Determining the right scaler is crucial for scaling your application based on the desired event source.
26+
27+
1. Visit the [KEDA Scalers documentation](https://keda.sh/docs/2.14/scalers/) and browse through the list of available scalers.
28+
2. Identify the scaler that matches the event source you want to use for scaling your application. For example:
29+
- If you want to scale based on incoming HTTP traffic, you would need the [HTTP Add-on](https://kedacore.github.io/http-add-on/).
30+
31+
> **Note:**
32+
> The HTTP Add-on is still in beta stage and may not provide the full functionality or stability expected in a production environment.
33+
34+
- If you want to scale based on messages in a RabbitMQ queue, you would need the **RabbitMQ scaler**.
35+
- If you want to scale based on a cron schedule, you would need the **Cron scaler**.
36+
3. Open the documentation page for your chosen scaler and familiarize yourself with its specific requirements and configuration options.
37+
38+
## Step 2: Install the Required Scaler (if needed)
39+
40+
Some scalers are part of the core KEDA installation, while others need to be installed separately as add-ons.
41+
42+
1. Refer to the documentation of your chosen scaler to check if it needs to be installed separately.
43+
2. If the scaler needs to be installed separately, follow the installation instructions provided in the scaler's documentation carefully.
44+
- The installation process typically involves running a command (e.g., `helm install` for Helm charts) or applying YAML manifests using `kubectl`.
45+
3. Verify that the scaler has been installed successfully by checking the output of the installation process or by running any provided verification commands.
46+
47+
## Step 3: Create a ScaledObject Configuration File
48+
49+
KEDA uses a custom resource called `ScaledObject` to define how your application should be scaled based on the chosen event source or trigger.
50+
51+
1. Create a new file (e.g., `scaledobject.yaml`) in a text editor or using the command line.
52+
2. Define the `ScaledObject` configuration in this file, following the structure and examples provided in the documentation of your chosen scaler.
53+
3. Typically, the `ScaledObject` configuration includes the following sections:
54+
- `metadata`: Specifies the name and namespace for the `ScaledObject`.
55+
- `spec.scaleTargetRef`: Identifies the Kubernetes deployment or other resource that should be scaled.
56+
- `spec.pollingInterval` (optional): Specifies how often KEDA should check for scaling events (defaults to 15 seconds).
57+
- `spec.cooldownPeriod` (optional): Specifies the cool-down period in seconds after a scaling event (defaults to 300 seconds).
58+
- `spec.maxReplicaCount` (optional): Specifies the maximum number of replicas to scale up to (defaults to 100).
59+
- `spec.triggers`: Defines the specific configuration for your chosen scaler, including any required parameters or settings.
60+
4. Refer to the scaler's documentation for detailed explanations and examples of the `triggers` section and any other required or optional configuration settings.
61+
5. Save the `scaledobject.yaml` file after making the necessary modifications.
62+
63+
## Step 4: Apply the ScaledObject Configuration
64+
65+
Once you have created the `ScaledObject` configuration file, apply it to your Kubernetes cluster using `kubectl`:
66+
67+
1. Open a terminal or command prompt and navigate to the directory containing the `scaledobject.yaml` file.
68+
2. Run the following command to apply the `ScaledObject` configuration:
69+
70+
```bash
71+
kubectl apply -f scaledobject.yaml
72+
```
73+
74+
```plaintext
75+
scaledobject.keda.sh/<scaled-object-name> created
76+
```
77+
78+
3. Verify that the `ScaledObject` has been created successfully by running:
79+
80+
```bash
81+
kubectl get scaledobjects
82+
```
83+
84+
This should display the `ScaledObject` you just created.
85+
86+
```plaintext
87+
NAME SCALETARGETKIND SCALETARGETNAME MIN MAX TRIGGERS AUTHENTICATION READY ACTIVE FALLBACK AGE
88+
<scaled-object-name> Deployment <deployment-name> 1 10 cpu <none> True False <none> 10s
89+
```
90+
91+
After applying the `ScaledObject` configuration, KEDA will start monitoring the specified event source and scale your application accordingly, based on the configurations you provided.
92+
93+
## Step 5: Monitor Scaling Events
94+
95+
You can monitor the scaling events and logs generated by KEDA using the following commands:
96+
97+
1. List all `ScaledObjects` in your cluster:
98+
99+
```bash
100+
kubectl get scaledobjects
101+
```
102+
103+
This will show you the current state of your `ScaledObject` and the number of replicas.
104+
105+
```plaintext
106+
NAME SCALETARGETKIND SCALETARGETNAME MIN MAX TRIGGERS AUTHENTICATION READY ACTIVE FALLBACK AGE
107+
<scaled-object-name> Deployment <deployment-name> 1 10 cpu <none> True False <none> 10s
108+
```
109+
110+
2. View the logs of the KEDA operator:
111+
112+
```bash
113+
kubectl logs -n keda -l app=keda-operator
114+
```
115+
116+
The KEDA operator logs will show you detailed information about scaling events, decisions made by KEDA based on the event source, and any errors or warnings.
117+
118+
```plaintext
119+
{"level":"info","ts":<timestamp>,"logger":"scalehandler","msg":"Successfully scaled deployment","scaledobject.Namespace":"<namespace>","scaledobject.Name":"<scaled-object-name>","scaler":<scaler-type>}
120+
```

‎content/docs/2.15/setupscaler.md

+120
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,120 @@
1+
+++
2+
title = "Setup Autoscaling with KEDA"
3+
weight = 500
4+
description = "Procedure to Setup a Scaler in KEDA"
5+
+++
6+
7+
## Prerequisites
8+
9+
1. **Kubernetes Cluster**:
10+
- Ensure you have a running Kubernetes cluster set up and accessible.
11+
- If you don't have a cluster yet, follow the [official Kubernetes documentation](https://kubernetes.io/docs/setup/) to create a new cluster suitable for your environment (local machine, cloud provider, etc.).
12+
13+
2. **KEDA Installation**:
14+
- KEDA needs to be installed on your Kubernetes cluster before you can use it.
15+
- Follow the [KEDA installation guide](https://keda.sh/docs/2.14/deploy/) carefully, including any prerequisites specific to your Kubernetes setup.
16+
- The installation guide provides instructions for different installation methods (e.g., YAML, Helm charts, etc.). Choose the method that suits your needs.
17+
18+
3. **kubectl**:
19+
- The `kubectl` command-line tool is required to interact with your Kubernetes cluster.
20+
- Follow the [official kubectl installation guide](https://kubernetes.io/docs/tasks/tools/#kubectl) to install `kubectl` on your operating system.
21+
- Once installed, configure `kubectl` to communicate with your Kubernetes cluster by following the cluster-specific instructions provided by your Kubernetes setup.
22+
23+
## Step 1: Identify the Scaler You Need
24+
25+
KEDA supports various scalers that correspond to different event sources or triggers. Determining the right scaler is crucial for scaling your application based on the desired event source.
26+
27+
1. Visit the [KEDA Scalers documentation](https://keda.sh/docs/2.14/scalers/) and browse through the list of available scalers.
28+
2. Identify the scaler that matches the event source you want to use for scaling your application. For example:
29+
- If you want to scale based on incoming HTTP traffic, you would need the [HTTP Add-on](https://kedacore.github.io/http-add-on/).
30+
31+
> **Note:**
32+
> The HTTP Add-on is still in beta stage and may not provide the full functionality or stability expected in a production environment.
33+
34+
- If you want to scale based on messages in a RabbitMQ queue, you would need the **RabbitMQ scaler**.
35+
- If you want to scale based on a cron schedule, you would need the **Cron scaler**.
36+
3. Open the documentation page for your chosen scaler and familiarize yourself with its specific requirements and configuration options.
37+
38+
## Step 2: Install the Required Scaler (if needed)
39+
40+
Some scalers are part of the core KEDA installation, while others need to be installed separately as add-ons.
41+
42+
1. Refer to the documentation of your chosen scaler to check if it needs to be installed separately.
43+
2. If the scaler needs to be installed separately, follow the installation instructions provided in the scaler's documentation carefully.
44+
- The installation process typically involves running a command (e.g., `helm install` for Helm charts) or applying YAML manifests using `kubectl`.
45+
3. Verify that the scaler has been installed successfully by checking the output of the installation process or by running any provided verification commands.
46+
47+
## Step 3: Create a ScaledObject Configuration File
48+
49+
KEDA uses a custom resource called `ScaledObject` to define how your application should be scaled based on the chosen event source or trigger.
50+
51+
1. Create a new file (e.g., `scaledobject.yaml`) in a text editor or using the command line.
52+
2. Define the `ScaledObject` configuration in this file, following the structure and examples provided in the documentation of your chosen scaler.
53+
3. Typically, the `ScaledObject` configuration includes the following sections:
54+
- `metadata`: Specifies the name and namespace for the `ScaledObject`.
55+
- `spec.scaleTargetRef`: Identifies the Kubernetes deployment or other resource that should be scaled.
56+
- `spec.pollingInterval` (optional): Specifies how often KEDA should check for scaling events (defaults to 15 seconds).
57+
- `spec.cooldownPeriod` (optional): Specifies the cool-down period in seconds after a scaling event (defaults to 300 seconds).
58+
- `spec.maxReplicaCount` (optional): Specifies the maximum number of replicas to scale up to (defaults to 100).
59+
- `spec.triggers`: Defines the specific configuration for your chosen scaler, including any required parameters or settings.
60+
4. Refer to the scaler's documentation for detailed explanations and examples of the `triggers` section and any other required or optional configuration settings.
61+
5. Save the `scaledobject.yaml` file after making the necessary modifications.
62+
63+
## Step 4: Apply the ScaledObject Configuration
64+
65+
Once you have created the `ScaledObject` configuration file, apply it to your Kubernetes cluster using `kubectl`:
66+
67+
1. Open a terminal or command prompt and navigate to the directory containing the `scaledobject.yaml` file.
68+
2. Run the following command to apply the `ScaledObject` configuration:
69+
70+
```bash
71+
kubectl apply -f scaledobject.yaml
72+
```
73+
74+
```plaintext
75+
scaledobject.keda.sh/<scaled-object-name> created
76+
```
77+
78+
3. Verify that the `ScaledObject` has been created successfully by running:
79+
80+
```bash
81+
kubectl get scaledobjects
82+
```
83+
84+
This should display the `ScaledObject` you just created.
85+
86+
```plaintext
87+
NAME SCALETARGETKIND SCALETARGETNAME MIN MAX TRIGGERS AUTHENTICATION READY ACTIVE FALLBACK AGE
88+
<scaled-object-name> Deployment <deployment-name> 1 10 cpu <none> True False <none> 10s
89+
```
90+
91+
After applying the `ScaledObject` configuration, KEDA will start monitoring the specified event source and scale your application accordingly, based on the configurations you provided.
92+
93+
## Step 5: Monitor Scaling Events
94+
95+
You can monitor the scaling events and logs generated by KEDA using the following commands:
96+
97+
1. List all `ScaledObjects` in your cluster:
98+
99+
```bash
100+
kubectl get scaledobjects
101+
```
102+
103+
This will show you the current state of your `ScaledObject` and the number of replicas.
104+
105+
```plaintext
106+
NAME SCALETARGETKIND SCALETARGETNAME MIN MAX TRIGGERS AUTHENTICATION READY ACTIVE FALLBACK AGE
107+
<scaled-object-name> Deployment <deployment-name> 1 10 cpu <none> True False <none> 10s
108+
```
109+
110+
2. View the logs of the KEDA operator:
111+
112+
```bash
113+
kubectl logs -n keda -l app=keda-operator
114+
```
115+
116+
The KEDA operator logs will show you detailed information about scaling events, decisions made by KEDA based on the event source, and any errors or warnings.
117+
118+
```plaintext
119+
{"level":"info","ts":<timestamp>,"logger":"scalehandler","msg":"Successfully scaled deployment","scaledobject.Namespace":"<namespace>","scaledobject.Name":"<scaled-object-name>","scaler":<scaler-type>}
120+
```

0 commit comments

Comments
 (0)