Vamp workflows are a convenient way to run Node JS based scripts that access the Vamp API to monitor and interact with running services. JavaScript workflows run in Vamp workflow agent containers ( - Vamp workflow agent) and are managed just like any other container inside your cluster, making them robust, scalable and dynamic. Workflows can be scheduled to run as a daemon, be triggered by Vamp events or to run at specified times.

Vamp ships with four default workflows:

  • Health - checks and stores the status of running services. Events stored by the health workflow are used by the Vamp UI.
  • Metrics - stores metrics on running services. Events stored by the metrics workflow are used by the Vamp UI.
  • Kibana - supports the easy creation of Kibana dashboards.
  • Allocation - calculates resource usage (CPU, memory)

System breeds

The following breeds are required by system workflows and should not be deleted: allocation, health, kibana, metrics, vamp-workflow-javascript

On this page:

Track running workflows

You can track executions of a running workflow in the Vamp UI. From the Workflows page, click on a workflow port to open the execution list. Click on a specific execution to open its log.

Create a workflow

Workflows can be created using Node JS based scripts running inside a Vamp workflow agent container, or use another language of preference - create an application or script that accesses the Vamp API and build it into a Docker container to be deployed by Vamp. Vamp tracks all revisions made to workflows and breeds, so you can check back and compare the current version against a previous version.

Try it yourself: Create a workflow that generates events

Workflow schedules

You can schedule a workflow to run as a daemon, be triggered by specific events or run according to a time schedule. See the examples for each below.

  • Scheduled as a daemon A workflow scheduled as a daemon will run continuously.
  schedule: daemon
  • Triggered by events
  event: # event with following tags will trigger the workflow
  - deployments:sava
  - cluster:runner

schedule:  # shortened notation in case of single event (still array can be used as above)
  event: archive:bluprints
  • Scheduled by time period, start (optional, by default starts now) and repeat (optional, by default runs forever). The time schedule period is in ISO8601 repeating interval notation. For example:
    period: P1Y2M3DT4H5M6S
    start: now # or e.g. start: 2016-12-03T08:15:30Z
    repeat: 10

Workflow dialects

You can use dialects to specify native commands for the underlying container platform in a workflow. Read more about Vamp dialects

JavaScript workflows

JavaScript workflows are executed by Vamp Workflow Agent ( - Vamp workflow agent). The system breed vamp-workflow-javascript is used to apply the standard enviroment variables, health checks and exposed ports. Instructions for using JavaScript to access the Vamp API can be found in the Vamp Node Client project ( - Vamp node client).

Example: Metrics workflow and breed

The below examples show a JavaScript workflow with a separately stored JavaScript breed. All additional environment variables, health checks, ports and scale will be added from the vamp-workflow-javascript breed.

name    : metrics
breed   : metrics
schedule: daemon

The JavaScript breed referenced in the metrics workflow will be executed by Vamp Workflow Agent ( - Vamp workflow agent). You could create a Javascript breed by sending the required JavaScript directly to the API to store as a breed:

  • Request syntax:
Content-Type: application/javascript
  • Request body:
'use strict';

let _ = require('highland');
let vamp = require('vamp-node-client');

let api = new vamp.Api();
let logger = new vamp.Log();
let metrics = new vamp.ElasticsearchMetrics(api);

let window = 30; // seconds

function publish(tags, metrics) {
  logger.log('metrics: [' + JSON.stringify(tags) + '] - ' + metrics);
  api.event(tags, metrics, 'metrics');

api.gateways().each(function (gateway) {

  metrics.average({ft: gateway.lookup_name}, 'Tt', window).each(function (response) {
    publish(['gateways:' +, 'gateway', 'metrics:rate'], response.rate);
    publish(['gateways:' +, 'gateway', 'metrics:responseTime'], response.average);

  api.namify(gateway.routes).each(function (route) {
    metrics.average({ft: route.lookup_name}, 'Tt', window).each(function (response) {
      publish(['gateways:' +, 'routes:' +, 'route', 'metrics:rate'], response.rate);
      publish(['gateways:' +, 'routes:' +, 'route', 'metrics:responseTime'], response.average);

What next?

Last updated on July 2, 2018