Alerts

PipelineWise sends failure alerts through handlers configured in config.yml. Replication and data-diff share the same routing and per-tap suppression.

Handlers

Available
Handler	Behaviour
Slack	Sends to the global channel and optional tap-specific channel.

Experimental
Handler	Limitation
VictorOps	Mock-endpoint coverage only; all taps use one routing key.

Slack

Create a Slack app with chat:write, install it to the workspace, and invite the bot to each destination channel.

alert_handlers:
  slack:
    token: "{{ env_var['SLACK_BOT_TOKEN'] }}"
    channel: "#pipeline-alerts"

Add a second destination for one tap:

slack_alert_channel: "#orders-alerts"

The tap-specific channel receives a copy; it does not replace the global channel.

VictorOps

Attention

Verify this handler against the real integration before relying on it for on-call response. It has only been exercised against a mocked endpoint.

alert_handlers:
  victorops:
    base_url: "https://alert.victorops.com/integrations/generic/.../alert"
    routing_key: "pipelinewise"

base_url must exclude the routing key. Requests time out after 10 seconds.

Routing and suppression

send_alert: false in a tap YAML suppresses its replication and data-diff alerts. It does not change command exit status or persisted run results.

Test every handler after configuration and monitor the handler itself. An alert channel is not a substitute for scheduler exit-status monitoring or target-data verification.