CNCF Atlas

English

日本語

English

日本語

Appearance

Getting Started ​

Verified against the source at commit 62b10c7 (18 commits past tag 0.10.0). Commands assume a Kubernetes cluster reachable through kubectl and a shell with curl.

Prerequisites ​

  • A Kubernetes cluster set as the current kubectl context. Drasi installs its components there.
  • curl (or wget) to fetch the CLI installer, which downloads a release binary from GitHub (cli/installers/install-drasi-cli.sh).
  • Network access to pull Drasi's container images during drasi init.

Install ​

Drasi is used through the drasi CLI. The repository ships an installer script that downloads the matching release binary into /usr/local/bin (cli/installers/install-drasi-cli.sh:15, install-drasi-cli.sh:117):

bash
curl -fsSL https://raw.githubusercontent.com/drasi-project/drasi-platform/main/cli/installers/install-drasi-cli.sh | /bin/bash

With the CLI in place, install the Drasi platform onto the cluster that is your current kubectl context. drasi init is the install command (cli/cmd/init.go):

bash
drasi init

A first working setup ​

The shortest working path is a Source, a Continuous Query, and a Reaction that shows the query output. The repository carries example resources under cli/ that match this shape (cli/test-source.yaml, cli/test-query.yaml).

  1. Apply a Source that connects to the system whose changes you want to observe. A PostgreSQL Source names the tables to watch (cli/test-source.yaml). drasi apply creates or updates resources (cli/cmd/apply.go).

    bash
    drasi apply -f source.yaml

  2. Apply a Continuous Query that defines what to observe. The query is Cypher, and it subscribes to the Source by id and returns the shape of the output (cli/test-query.yaml):

    yaml
    kind: ContinuousQuery
apiVersion: v1
name: query1
spec:
  mode: query
  sources:
    subscriptions:
      - id: foo
  query: >
    MATCH
      (i:Item {Category: '1'})
    RETURN
      i.ItemId AS Id,
      i.Name as Name,
      i.Category as Category
    bash
    drasi apply -f query.yaml

  3. Apply a Reaction to observe the output. The tutorial uses a debug Reaction that displays the rows a query adds, updates, and deletes.

    bash
    drasi apply -f reaction.yaml

Verify it works ​

List the resources Drasi is managing and confirm the query is running:

bash
drasi list query

Then describe the query to see its status and current result set (cli/cmd/describe.go):

bash
drasi describe query query1

Change a row in the source table that matches the query pattern (here, an Item with Category 1), and the debug Reaction shows the corresponding added, updated, or deleted result. A green status on drasi describe means the Source is bootstrapped and the query is evaluating live changes.

Where to go next ​

The official Getting Started tutorial walks the full Source, Continuous Query, and debug Reaction flow on Kubernetes at https://drasi.io/drasi-kubernetes/getting-started/, and the documentation site at https://drasi.io/ covers the prebuilt Sources and Reactions, the Continuous Query language, and how-to guides for production concerns. For custom integrations, the Source and Reaction SDKs live under sources/sdk and reactions/sdk in the repository.