windmill-labs/windmill · error

Timed out waiting for the project to become reachable

Error message

Timed out waiting for the project to become reachable

What it means

waitUntilSupabaseHealthy polls the Supabase projects list until the newly created project reports status ACTIVE_HEALTHY. If the loop exhausts its attempts before the project becomes reachable, it throws this timeout error. Newly created Supabase projects usually take a few minutes to provision, so this fires when provisioning is slower than the wait budget or the project failed to start.

Source

Thrown at frontend/src/lib/components/workspaceSettings/supabaseProvisioning.ts:177

	onStatus?: (status: string | undefined) => void,
	attempts = 60
): Promise<SupabaseProject> {
	for (let i = 0; i < attempts; i++) {
		await new Promise((r) => setTimeout(r, 5000))
		let list: SupabaseProject[]
		try {
			list = await listSupabaseProjects(token)
		} catch (err) {
			// A transient failure is worth another poll; an expired token is not -- retrying it
			// burns five minutes and then reports a timeout, which names the wrong problem.
			if (!get(oauthStore)?.access_token) throw err
			continue
		}
		const project = list?.find?.((p) => projectRef(p) === projectId)
		if (project?.status === 'ACTIVE_HEALTHY') return project
		onStatus?.(project?.status)
	}
	throw new Error('Timed out waiting for the project to become reachable')
}

/**
 * The session-mode Supavisor endpoint of the project's primary database.
 *
 * Which pooler a project sits behind is assigned by Supabase, not derived from its
 * region: constructing `aws-0-<region>.pooler.supabase.com` is wrong for every project
 * that landed on another one, and the resulting resource never connects.
 */
export async function getSupabasePooler(token: string, projectId: string): Promise<SupabasePooler> {
	const res = await fetch(`${base}/api/oauth/get_supabase_pooler/${projectId}`, {
		headers: headers(token)
	})
	const configs: SupabasePooler[] = await unwrap(res, 'Could not read the connection details')
	const primary = configs.filter((c) => c.database_type === 'PRIMARY')
	const pooler = primary.find((c) => c.pool_mode === 'session') ?? primary[0] ?? configs[0]
	if (!pooler) throw new Error('Supabase returned no connection details for this project')
	return pooler

View on GitHub (pinned to e474e8803c)

Solutions

1. Wait a few minutes and re-run the setup/provisioning step — the project usually becomes ACTIVE_HEALTHY on its own.
2. Check the project's status in the Supabase dashboard; if it's failed/stuck, delete it and create a new one.
3. Verify the projectId being polled matches the project returned by the create call.
4. Increase the polling timeout/attempts in supabaseProvisioning.ts if provisioning in your region is consistently slow.

const project = (await listSupabaseProjects(token)).find((p) => p.id === projectId)
if (project && project.status !== 'ACTIVE_HEALTHY') console.log('still', project.status, '- keep waiting')

function isHealthy(p?: { status?: string }): p is { status: 'ACTIVE_HEALTHY' } & Record<string, unknown> {
  return !!p && p.status === 'ACTIVE_HEALTHY'
}

try {
  await waitUntilSupabaseHealthy(token, projectId, onStatus)
} catch (e) {
  if (e.message.includes('Timed out')) {
    // resume provisioning later; project may still be coming up
  } else throw e
}

When it happens

Trigger: runSetup creates a Supabase project and polls for health; the project stays in another status (e.g. 'COMING_UP', 'RESTORING') or disappears from the list for the whole polling window, then the timeout is thrown.

Common situations: Supabase provisioning backlogs making new projects take longer than the wait window; project stuck in an error state; wrong projectId being matched (created project ref doesn't match the one polled); heavy region with slow cold-start.

Understand the failure class

• Timeouts: ETIMEDOUT, deadlines, and hung requests — what actually expires when a request times out.

Related errors

• Timed out waiting for flow ${id} to complete
• Timed out waiting for job ${id}
• Timed out waiting for job ${id} to complete
• Giving up polling job ${jobId} after ${MAX_CONSECUTIVE_POLL_
• Failed to poll dependencies job ${jobId}: ${e?.message ?? e}

AI-assisted analysis of windmill-labs/windmill@e474e8803c (2026-09-03). Data as JSON: /api/errors/38bd14b4e2950973. Report an issue: GitHub.