Deploy to cloud¶

Run your flows on Repster Cloud: your schedules fire without your machine on, and each run gets its own isolated worker, created on demand. The pipeline code is exactly what you ran locally.

Prerequisites¶

A working local flow (see Quickstart)
A Repster account (you'll create one during rp login if you don't have one yet)

1. Authenticate¶

rp login

Repster walks you through account creation or login interactively. Your credentials are saved to ~/.repster/config.toml. You never need to run rp login before deploying; any cloud command that needs auth will walk you through it inline.

2. Push secrets¶

If your pipeline uses environment variables (API keys, database credentials), push them before deploying:

rp secrets push

This reads your .env file and stores the values, encrypted, in the Repster Cloud vault. See Manage secrets for details.

3. Deploy¶

rp deploy

Repster packages your project (source files + uv.lock), uploads it, and registers your flows. Deployments are immutable; see How Repster works.

4. Trigger a cloud run¶

rp run --cloud hackernews_top

The run is dispatched to an isolated cloud worker. You get the run ID immediately; the flow executes asynchronously.

5. Monitor the run¶

rp status                         # last N runs (local + cloud, combined)
rp logs --run <run-id>            # full log output

6. View in the UI¶

Log in to the Repster Cloud UI (repster.online) to see runs, logs, deployments, and secrets in one place.

Re-deploying¶

After changing your pipeline code:

rp deploy        # upload new version
rp run --cloud hackernews_top   # runs the latest deployment

Old deployments are preserved; runs always reference the deployment they were dispatched against.