Pipeline watch agent Live in lab

Guards every change before it bites.

Most pipeline incidents start with a change. When a connector configuration or a schema changes, or an application that writes to your source ships a new release, the watch agent follows the pipeline for an hour and flags drift against its own baseline, while the change is fresh.

Pipeline watchLive in lab
MON-0105On-call engineer

"Watch every connector change for two hours and flag heartbeat drift."

Orders connector watchActive
Trigger
Connector config change
Watches
Heartbeat age, unconfirmed WAL, task restarts, pipeline health
Window
2 h, against the pipeline's own 7-day baseline

Change-triggered

Starts a watch when the connector configuration or the topic's schema changes, or when an application that writes to your source deploys a new release.

Baseline, not guesswork

Compares heartbeat age, unconfirmed WAL and how often the connector's tasks restart with the pipeline's own last 7 days.

Flags early

Drift is told once, while the change is fresh, with the evidence and the change that started the watch.

Set in a sentence

Ask PelQi to watch the pipeline for two hours, or press Start a watch in the console.

Live in lab What it does today

  • Watches started by connector configuration and schema changes, and by new releases of the applications that write to your source
  • Baselines for heartbeat age, unconfirmed WAL and task restarts from the last 7 days
  • Watches requested in plain language through PelQi, or with one button
  • A downstream reader that stops or falls behind counts as drift
  • Every watch kept with its verdict: drift seen or no drift
Early access

Give every change a watch

Varqeni is in early access. Tell us about your pipelines and we will set up a trial around them.