A Zeebe community extension that provides a GraphQL query API over Zeebe's data. The data is imported from the broker using the Hazelcast exporter and aggregated in the service.
The docker image is published to GitHub Packages .
Run the following command to start the application:
docker run -p 9000:9000 ghcr.io/camunda-community-hub/zeeqs:latest
- Ensure that a Zeebe broker is running with a Hazelcast exporter
- Forward the Hazelcast port to the docker container (default:
5701) - Configure the connection to Hazelcast (default:
localhost:5701)ZEEBE_CLIENT_WORKER_HAZELCAST_CONNECTION(environment variable)zeebe.client.worker.hazelcast.connection(application.yaml)
Or, if you run it on your local machine (Linux only):
docker run --network="host" ghcr.io/camunda-community-hub/zeeqs:latest
After the application is started, the GraphQL endpoint is available under http://localhost:9000/graphql.
Go to http://localhost:9000/graphiql and explore the GraphQL API using GraphiQL.
For a local setup, the repository contains a docker-compose file. It contains multiple profiles for different configurations.
Use an in-memory database (H2):
docker-compose --profile in-memory up
Use a PostgreSQL database:
docker-compose --profile postgres up
After the application is started, the GraphQL endpoint is available under http://localhost:9000/graphql.
Go to http://localhost:9000/graphiql and explore the GraphQL API using GraphiQL.
By default, the API endpoint is available under the port 9000. And the database is only in-memory (i.e. not
persistent).
To configure the application, look at the application.yml and the docker-compose file. See here on how to configure a Spring Boot application in general.
The application provides a GraphQL endpoint under /graphql.
You can query the API via HTTP POST request and a JSON body containing the query.
For example:
curl \
-X POST \
-H "Content-Type: application/json" \
--data '{ "query": "{ processes { nodes { key } } }" }' \
http://localhost:9000/graphql
The best way to explore and inspect the GraphQL API is to use the integrated web-based tool GraphiQL. It is available under http://localhost:9000/graphiql.
The GraphQL API provides the following queries:
Example query
Query all processes including their number of process instances:
{
processes {
totalCount
nodes {
key
bpmnProcessId
version
processInstances {
totalCount
}
}
}
}Response:
{
"data": {
"processes": {
"totalCount": 3,
"nodes": [
{
"key": "2251799813685249",
"bpmnProcessId": "demo-process",
"version": 1,
"processInstances": {
"totalCount": 3
}
},
{
"key": "2251799813685251",
"bpmnProcessId": "demo-2",
"version": 1,
"processInstances": {
"totalCount": 2
}
},
{
"key": "2251799813685269",
"bpmnProcessId": "demo-3",
"version": 1,
"processInstances": {
"totalCount": 1
}
}
]
}
}
}In order to limit the response size and the query processing, each list query uses pagination to return only a subset of the data. By default, it returns the first 10 items of the list.
In addition to the items, the query result contains also the total count of the items.
{
processes(perPage: 10, page: 0) {
totalCount
nodes {
key
}
}
}Some queries allow to filter the result by passing arguments in the query.
{
processInstances(stateIn: [ACTIVATED]) {
nodes {
key
}
}
}The GraphQL API provides the following subscriptions:
The subscriptions are exposed via WebSockets over the same endpoint /graphql.
Example subscription
Subscribe to any updates of process instances:
subscription {
processInstanceUpdates {
processInstance {
key
state
}
updateType
}
}Response:
{
"data": {
"processInstanceUpdates": {
"processInstance": {
"key": "2251799813685251",
"state": "ACTIVATED"
},
"updateType": "ELEMENT_INSTANCE"
}
}
}Build with Maven
mvn clean install
This project adheres to the Contributor Covenant Code of Conduct. By participating, you are expected to uphold this code. Read more about the Camunda Community Code of Conduct and how to report unacceptable behavior.