Watch it run. On the graph you built.
Nodes light up as the run goes: pending, running, done, failed. Every cluster you enrolled, on one screen, for everyone you let in.
One UI per cluster, and its own set of credentials.
The Argo Workflows UI serves the cluster it was installed on. There is no view across all of them, so prod, staging and two regions is four addresses to keep straight.
Signing somebody in is the easy half. What they can see once they are in is Kubernetes RBAC. Showing one team one workflow means Roles and bindings on the cluster, and a namespace is as narrow as it goes.
Live status, node by node
Status from the cluster lands on the node it belongs to, while the run is still going.
- A failed step centers the canvas and shows its error.
- Rerun a finished run with the same parameters.
- Email or Slack the moment a run fails. The same switch covers a connector going offline.
Every cluster on one screen
Prod, staging and every region land in the same run list, and each trigger names the cluster it fires at.
- People sign in to dagweave once, rather than collecting a kubeconfig per cluster.
- Who sees what is a grant you make here, down to a single workflow. No Role on a cluster to edit.
Every trigger lands on the same canvas
A run started by a webhook overnight shows up exactly like one you clicked.
- A workflow that takes an input file runs on one you pick in the browser. No bucket, no key, no aws s3 cp first.
manual -> run it now, from the canvas
file -> pick one in the browser, it runs on that
cron -> scheduled in dagweave, or deployed as a CronWorkflow
webhook -> shared secret, or HMAC + nonce
github -> push / pull_request / tag
sns -> a topic, signature verifiedWatch your next run on the canvas.
dagweave is in early access. Leave an email and we will bring you on as we open it up.