Configuration via LCP.json
Each service in your Liferay Cloud environments has an LCP.json
file that you can use to configure the service. You can configure properties like the service ID, memory, number of CPUs, environment variables, volumes, and much more.
This table lists and describes the properties you can add in LCP.json
:
Field | Type | Default Value | Description |
---|---|---|---|
id | String | random | The service ID |
image | String | "" | The service image from Docker Hub |
env | Object | undefined | Environment variables |
loadBalancer | Object | {} | Declaration of exposed ports and domains |
cpu | Number | 1 | Number of CPUs |
scale | Number | 1 | Starting number of instances |
memory | Number | 512 | Amount of memory (MB) |
volumes | Object | undefined | Paths to persist data |
readinessProbe | Object | { "timeoutSeconds": 5 } | Service readiness check |
livenessProbe | Object | { "timeoutSeconds": 5 } | Service liveness check |
dependencies | Array | [] | Dependency deployment order |
kind | String | Deployment | Deployment type (e.g, Deployment or StatefulSet) |
strategy | Object | { "type": "RollingUpdate" } | Deployment strategy (e.g, RollingUpdate or Recreate) |
ports | Array | [] | Declaration of ports and protocols |
environments | Object | {} | Environment-specific configurations |
deploy | Boolean | true | Whether the service will be deployed for the specified environment. Only use this property inside the environments property; not at the root level. See the sample LCP.json file. |
autoscale | Object | { "cpu": 80, "memory": 80 } | The target average utilization for CPU and memory in auto-scaling. For more information about how this works with auto-scaling, see Auto-scaling. |
Usage
Here’s an example LCP.json
file that uses all the properties:
{
"id": "myservice",
"image": "liferaycloud/example",
"env": {
"DB_USER": "root",
"DB_PASSWORD": "pass123"
},
"loadBalancer": {
"cdn": true,
"targetPort": 3000,
"customDomains": ["example.com"],
"ssl": {
"key": "...",
"crt": "..."
}
},
"cpu": 2,
"scale": 2,
"memory": 2048,
"volumes": {
"storage": "/opt/storage"
},
"livenessProbe": {
"timeoutSeconds": 5,
"httpGet": {
"path": "/status",
"port": 3000
},
"initialDelaySeconds": 40,
"periodSeconds": 5,
"successThreshold": 5
},
"readinessProbe": {
"timeoutSeconds": 5,
"exec": {
"command": ["cat", "/tmp/healthy"]
},
"initialDelaySeconds": 40,
"periodSeconds": 5
},
"dependencies": ["service1", "service2"],
"kind": "StatefulSet",
"strategy": {
"type": "RollingUpdate"
},
"ports": [
{
"port": 3400,
"targetPort": 7000,
"protocol": "TCP"
},
{
"port": 9000,
"targetPort": 8000,
"protocol": "TCP",
"external": true
}
],
"environments": {
"prd": {
"memory": 4096,
"cpu": 6
},
"dev": {
"deploy": false
}
},
"autoscale": {
"cpu": 90,
"memory": 90
}
}
The environments
Configuration
The environments
configuration allows for environment-specific overrides to the LCP.json
configuration.
For example, the CI service should only be deployed to the infra
environment, so this environments
configuration overrides the behavior ("deploy": false
) specifically for the infra
environment:
{
"id": "ci",
"memory": 8192,
"cpu": 4,
"deploy": false,
"environments": {
"infra": {
"deploy": true
}
}
}
Overriding JSON Objects
For JSON object configurations (such as a loadBalancer
configuration), specify only the specific fields that you want to override per environment. Any other fields are inherited from the existing configuration by default.
In this example, the environments
configuration disables the CDN for the dev
environment, but keeps the target port of 80
for all environments.
{
"id": "webserver",
"memory": 512,
"cpu": 2,
"loadBalancer": {
"targetPort": 80,
"cdn": true
},
"environments": {
"dev": {
"loadBalancer": {
"cdn": false
}
}
}
}
Overriding JSON Arrays
Unlike JSON objects, if you override a JSON array in an environments
configuration, you must define the whole array in the override. Any elements of the array that you don’t include are not included in the environment-specific result.
In this example overriding the ports
array, the environments
configuration configures the uat
environment with only one port (but two for all other environments).
{
"id": "database",
"memory": 1024,
"cpu": 2,
"ports": [
{
"port": 3306,
"external": false
},
{
"port": 3000,
"external": false
}
],
"environments": {
"uat": {
"ports": [
{
"port": 3306,
"external": true
}
]
}
}
}