zeebe-cloud-canary
v1.1.4
Published
A Zeebe worker that monitors the aliveness connection with the Zeebe cluster
Downloads
12
Readme
Zeebe Cloud Canary
Use this package to allow your Zeebe applications to alert when they are not able to receive work from the broker - whether due to a broker outage or a dropped connection. Read more about the use-cases on the Zeebe blog.
The ZeebeCanary
class deploys a workflow to the broker that starts a loop with a user-configured heartbeat interval, and spawns a worker that services the loop.
Whenever the canary worker services the loop task, it pings the "chirp" url.
If the canary does not get a chirp task in 1.5x the heartbeat interval, it can ping a "squawk" url.
The worker starts the next loop iteration as a new workflow instance, so that you are not running a non-reapable long-running process (that would mess with your snapshot size on disk and recovery time).
The canary uses micromustache to template the ZeebeCanary.CanaryId
into the workflow before it is deployed, to allow you to have canaries for multiple applications / instances running on the same broker.
Installation
npm i zeebe-cloud-canary
Use
import { ZeebeCanary } from 'zeebe-cloud-canary';
// Uses the zeebe-node zero-conf constructor, either localhost or from ENV
const canary = new ZeebeCanary({
ChirpUrl: `${healthchecks_io_url}`,
CanaryId: 'some-application-id',
HeartbeatPeriodSeconds: 300
})
You can pass in the ZBClient constructor options explicitly, like this:
import { ZeebeCanary } from 'zeebe-cloud-canary';
// Uses an explicit ZBClient constructor for configuration
const canary = new ZeebeCanary({
ChirpUrl: '',
CanaryId: 'some-application-id',
HeartbeatPeriodSeconds: 300,
ZBConfig: {
hostname: `${zeebe_address}:${optional_port}`,
longPoll: 55000
}
})
If you use an external healthcheck ping like healthchecks.io for the ChirpUrl
, it will squawk if the canary does not chirp on time, and if the canary process dies, so it will also alert you on client application failure.
You can use your own SquawkUrl
, and it will be pinged if the canary does not get a chirp task in 1.5x the heartbeat interval - but not if the canary process itself (your application) dies.
Running in Docker
You can run a fully dockerised Zeebe Cloud Canary like this:
ZEEBE_ADDRESS=<some_url_reachable_from_docker> docker-compose up
By default the canary chirps and squawks in the container log.
You can set the following environment variables:
CANARY_HEARTBEAT_SEC
CHIRP_URL
DEBUG
SQUAWK_URL
ZEEBE_ADDRESS
ZEEBE_CLIENT_ID
ZEEBE_CLIENT_SECRET