KEDA

https://github.com/kedacore/keda
Admin
KEDA enables fine-grained autoscaling for Kubernetes event-driven workloads with support for scaling...
Tokens:276,493
Snippets:1,643
Trust Score:8.9
License:Apache-2.0
Update:1 week ago
Show doc for...
Tokens:
Raw
### Customize test setup with custom images

Source: https://github.com/kedacore/keda/blob/main/tests/README.md

Example config to customize the test setup, e.g., deploying KEDA using custom images. If imageRegistry and imageRepo are empty, the default KEDA repository and registry from the Makefile are used.

```yaml
keda:
  # These are the default values.
  skipSetup: false # If true, the test script will skip the setup phase.
  skipCleanup: false # If true, the test script will skip the cleanup phase.
  imageRegistry: ""
  imageRepo: ""
```

--------------------------------

### Example config file with includes

Source: https://github.com/kedacore/keda/blob/main/tests/README.md

Example YAML config that only runs the aws_cloudwatch, cpu, and kafka scaler tests. Categories not specified will not run.

```yaml
testCategories:
  # This will only run the aws_cloudwatch, cpu, and kafka scaler tests, and exclude the rest.
  scalers:
    mode: include
    tests:
      - aws/aws_cloudwatch # you can also specify a deeper nested test by it's "directory path"
      - cpu
      - kafka
  # Since the other categories are not specified, all tests in those categories will not be run.
```

--------------------------------

### Install @codama/dynamic-codecs with pnpm

Source: https://github.com/kedacore/keda/blob/main/CONTRIBUTING.md

Install the package using pnpm. Note that this package is not included in the main codama package.

```sh
pnpm install @codama/dynamic-codecs
```

--------------------------------

### GetMetricSpecForScaling implementation example

Source: https://github.com/kedacore/keda/blob/main/CREATE-NEW-SCALER.md

Example of implementing GetMetricSpecForScaling for a scaler. It creates an ExternalMetricSource with a metric name generated using GenerateMetricNameWithIndex and a target value from GetMetricTarget. The metric name must start with 'sX-' where X is the triggerIndex.

```golang
func (s *artemisScaler) GetMetricSpecForScaling() []v2.MetricSpec {
	externalMetric := &v2.ExternalMetricSource{
		Metric: v2.MetricIdentifier{
			Name: GenerateMetricNameWithIndex(s.metadata.triggerIndex, kedautil.NormalizeString(fmt.Sprintf("%s-%s-%s", "artemis", s.metadata.brokerName, s.metadata.queueName))),
		},
		Target: GetMetricTarget(s.metricType, s.metadata.queueLength),
	}
	metricSpec := v2.MetricSpec{External: externalMetric, Type: artemisMetricType}
	return []v2.MetricSpec{metricSpec}
}
```

--------------------------------

### Example config file with exclusions

Source: https://github.com/kedacore/keda/blob/main/tests/README.md

Example YAML config that excludes the cpu test from scalers and excludes all tests in other categories. Omitting an existing category runs all tests in that category.

```yaml
testCategories:
  # This will run all tests in the scalers category, but exclude the cpu test.
  scalers:
    mode: exclude
    tests:
      - cpu
  # Using mode: exclude and omitting the tests list will exclude all tests in the category.
  internals:
    mode: exclude
  secret-providers:
    mode: exclude
  sequential:
    mode: exclude
```

--------------------------------

### Run all e2e tests

Source: https://github.com/kedacore/keda/blob/main/tests/README.md

Run the full e2e test suite. The setup test must be run once before the scaler tests, and the cleanup test can be skipped if you want to keep testing.

```bash
go test -v -tags e2e ./utils/setup_test.go        # Only needs to be run once.
go test -v -tags e2e ./scalers/...
go test -v -tags e2e ./utils/cleanup_test.go      # Skip if you want to keep testing.
```

--------------------------------

### Run a specific e2e test

Source: https://github.com/kedacore/keda/blob/main/tests/README.md

Run a single scaler test file. Assumes the setup has been run before.

```bash
go test -v -tags e2e ./scalers/azure/azure_queue/azure_queue_test.go # Assumes that setup has been run before
```

--------------------------------

### Logging time encoding values and output

Source: https://github.com/kedacore/keda/blob/main/BUILD.md

Example of some of the logging time encoding values and their corresponding output. The values are epoch, iso8601, rfc3339, and rfc3339nano.

```text
epoch - 1.6533943565181081e+09
iso8601 - 2022-05-24T12:10:19.411Z
rfc3339 - 2022-05-24T12:07:40Z
rfc3339nano - 2022-05-24T12:10:19.411Z
```

--------------------------------

### Resolve link nodes with getNodeCodec

Source: https://github.com/kedacore/keda/blob/main/CONTRIBUTING.md

Example showing how link nodes are resolved. The codec for a linked defined type is resolved using the target type.

```ts
// Here we define a program with two types, one of which is a link to the other.
const root = rootNode(
    programNode({
        definedTypes: [
            definedTypeNode({ name: 'slot', type: numberTypeNode('u64') }),
            definedTypeNode({ name: 'lastSlot', type: definedTypeLinkNode('slot') }),
        ],
        name: 'myProgram',
        publicKey: '1111',
    }),
);

// The codec for the linked `lastSlot` defined type is resolved using the `slot` defined type.
const codec = getNodeCodec([root, root.program, root.program.definedTypes[1]]);
expect(codec.encode(42)).toStrictEqual(hex('2a00000000000000'));
expect(codec.decode(hex('2a00000000000000'))).toBe(42n);
```

--------------------------------

### Get codec for a node using getNodeCodec

Source: https://github.com/kedacore/keda/blob/main/CONTRIBUTING.md

Retrieve a codec for a node by providing its full NodePath. The codec can then encode and decode data.

```ts
const codec = getNodeCodec([root, program, definedType]);
const bytes = codec.encode(someData);
const decodedData = codec.decode(bytes);
```

--------------------------------

### Message Flow Sequence

Source: https://github.com/kedacore/keda/blob/main/tests/README.md

Shows the message flow between parent and daemon processes, including task start, execution, coordinator checks, and task stop events.

```text
parent                                daemon
  |-- task:start {path,cron,options} -->|
  |                                     | import(taskPath); new InlineScheduledTask; start()
  |<-- daemon:started ------------------|
  |<-- task:started {context} ----------|
  |<-- execution:started {context} -----|
  |-- coordinator:shouldRun {key,reqId} |
  |              (outbound; parent runs real coordinator)
  |<-- coordinator:result {reqId,allowed} |
  |<-- execution:finished {context,jsonError?} |
  |-- task:stop ----------------------->|
  |<-- task:stopped --------------------|
  |<-- [daemon exits; parent: killForkWhenSettled]
```

--------------------------------

### Deploy KEDA with make deploy

Source: https://github.com/kedacore/keda/blob/main/BUILD.md

Deploys CRDs and KEDA into the 'keda' namespace. Run after setting up certificates and CRDs.

```bash
make deploy
```

--------------------------------

### VS Code launch.json for KEDA metrics server

Source: https://github.com/kedacore/keda/blob/main/BUILD.md

Configuration for debugging the KEDA metrics server in VS Code. Includes authentication and authorization flags. Replace PATH_TO_YOUR_KUBECONFIG with your actual path.

```json
{
    "configurations": [
        {
            "name": "Launch metrics-server",
            "type": "go",
            "request": "launch",
            "mode": "auto",
            "program": "${workspaceFolder}/cmd/adapter/main.go",
            "env": {
                "WATCH_NAMESPACE": "",
                "KEDA_CLUSTER_OBJECT_NAMESPACE": "keda"
            },
            "args": [
                "--authentication-kubeconfig=PATH_TO_YOUR_KUBECONFIG",
                "--authentication-skip-lookup",
                "--authorization-kubeconfig=PATH_TO_YOUR_KUBECONFIG",
                "--lister-kubeconfig=PATH_TO_YOUR_KUBECONFIG",
                "--secure-port=6443",
                "--v=5"
            ],
        }
    ]
   }
```

--------------------------------

### Generate self-signed keystore with keytool

Source: https://github.com/kedacore/keda/blob/main/tests/scalers/pulsar/README.md

Creates a self-signed keystore with a long expiration (36500 days) and includes Subject Alternative Names for the two test services. The password is 'protected'.

```shell
keytool \
  -keystore server.jks  -storepass protected  -deststoretype pkcs12 \
  -genkeypair -keyalg RSA -validity 36500 \
  -dname "CN=pulsar.apache.org,O=pulsar,OU=pulsar" \
  -ext "SAN=DNS:pulsar-partitioned-topic-test.pulsar-partitioned-topic-test,DNS:pulsar-non-partitioned-topic-test.pulsar-non-partitioned-topic-test"
```

--------------------------------

### Scale down keda-operator deployment

Source: https://github.com/kedacore/keda/blob/main/BUILD.md

Scales the keda-operator Deployment to 0 replicas in the 'keda' namespace to allow local debugging.

```bash
kubectl scale deployment/keda-operator --replicas=0 -n keda
```

--------------------------------

### Base64 encode certificate and key for secret

Source: https://github.com/kedacore/keda/blob/main/tests/scalers/pulsar/README.md

Encodes the certificate and key files to base64 for placement in a Kubernetes secret. These commands are for MacOS and copy the output to the clipboard.

```shell
cat servercert.pem | base64 | pbcopy
```

```shell
cat serverkey.pem | base64 | pbcopy
```

--------------------------------

### Extract private key with openssl

Source: https://github.com/kedacore/keda/blob/main/tests/scalers/pulsar/README.md

Extracts the private key from the keystore into a PEM file. Requires entering the password 'protected'.

```shell
openssl pkcs12 -in server.jks -nodes -nocerts -out serverkey.pem
```

--------------------------------

### Valid config with empty testCategories

Source: https://github.com/kedacore/keda/blob/main/tests/README.md

A valid config file with testCategories set to an empty map. This runs all tests in all categories.

```yaml
# This will run all tests in all categories.
testCategories: {}
```

--------------------------------

### Extract public key with openssl

Source: https://github.com/kedacore/keda/blob/main/tests/scalers/pulsar/README.md

Extracts the public certificate from the keystore into a PEM file. Requires entering the password 'protected'.

```shell
openssl pkcs12 -in server.jks -nokeys -out servercert.pem
```

--------------------------------

### Run e2e tests with a custom config file

Source: https://github.com/kedacore/keda/blob/main/tests/README.md

Use the E2E_TEST_CONFIG environment variable to run a subset of tests with a custom config file. Environment variables override config file fields.

```bash
E2E_TEST_CONFIG=tests/example-config.yaml go test -v -tags e2e ./tests/run-all.go
```

--------------------------------

### VS Code launch.json for KEDA operator

Source: https://github.com/kedacore/keda/blob/main/BUILD.md

Configuration for debugging the KEDA operator in VS Code. Sets WATCH_NAMESPACE to empty and KEDA_CLUSTER_OBJECT_NAMESPACE to 'keda'. Place in .vscode/launch.json.

```json
{
    "configurations": [
         {
            "name": "Launch operator",
            "type": "go",
            "request": "launch",
            "mode": "debug",
            "program": "${workspaceFolder}/cmd/operator/main.go",
            "env": {
                "WATCH_NAMESPACE": "",
                "KEDA_CLUSTER_OBJECT_NAMESPACE": "keda"
            }
        }
    ]
   }
```

--------------------------------

### Config with dryRun enabled

Source: https://github.com/kedacore/keda/blob/main/tests/README.md

Set dryRun to true to check what tests would be executed without actually running them.

```yaml
dryRun: true
keda:
  # ...omitted
testCategories:
  # ...omitted
```

--------------------------------

### Invalid config missing testCategories

Source: https://github.com/kedacore/keda/blob/main/tests/README.md

This config is invalid and will fail with an error message about testCategories being a required field.

```yaml
testCategories:
# Error loading test config: testCategories is a required field. Did you mean to set this to an empty map?
```

--------------------------------

### Grant anonymous access to external metrics API

Source: https://github.com/kedacore/keda/blob/main/BUILD.md

ClusterRoleBinding manifest that grants system:anonymous access to the external metrics API. Warning: This is potentially unsecure and should not be used on production clusters.

```yaml
apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRoleBinding
metadata:
   name: grant-anonymous-access-to-external-metrics
roleRef:
   apiGroup: rbac.authorization.k8s.io
   kind: ClusterRole
   name: keda-external-metrics-reader
subjects:
- kind: User
  name: system:anonymous
  namespace: default
```