[nishantmonu51]

• Since Move resource checks to deployment reconciliation #9653, ValidateDeploymentsWorker schedules ReconcileDeployment for every prod deployment twice an hour. For a running deployment the job always called UpdateDeploymentStatus(Running, ""), which sets updated_on = now() even when nothing changed.
• The runtime's configReloader uses updated_on as its "deployment changed" signal, so every reload forced a repo pull and a full re-parse of the project (a controller restart for archive-backed projects), cancelling in-flight model reconciles.
• The job now re-fetches the deployment and only calls UpdateDeploymentStatus when the status or status message actually differs. The re-fetch is needed because UpdateDeploymentInner may have moved the status to Updating in the DB without updating the job's in-memory copy.
• Real changes still propagate: UpdateDeployment, UpdateDeploymentDesiredStatus and UpdateDeploymentSafe bump updated_on before the job runs, and the runtime compares variables directly.
• Behavior change: the orphaned prod deployment teardown in ValidateDeploymentsWorker (non-primary prod deployment, not stopped, updated_on older than 3h) can fire again, as it did before v0.89. It could never trigger for a running deployment while updated_on was refreshed every 30 minutes.
• TestRuntimeDeployments now asserts that two no-op reconciles leave updated_on and the project parser's state version unchanged, and that a reconcile after a --prod-slots change leaves the deployment Running.

Steps to reproduce:

1. Deploy a project to Rill Cloud and wait for its prod deployment to reach RUNNING.
2. Watch deployments.updated_on for that deployment and the runtime's parser logs for its instance.
3. Make no changes to the project, its environment variables or its deployment for an hour.
4. Expected: updated_on stays constant and the project is not re-parsed.
   Actual: updated_on moves at every */30 run of ValidateDeploymentsWorker, and the runtime re-parses the whole project on each of its following config reloads.

Checklist:

• Covered by tests
• Ran it and it works as intended
• Reviewed the diff before requesting a review
• Checked for unhandled edge cases
• Linked the issues it closes
• Checked if the docs need to be updated. If so, create a separate Linear DOCS issue
• Intend to cherry-pick into the release branch
• I'm proud of this work!

---

Developed in collaboration with Claude Code

[k-anshul]

This doesn't seem safe on a high level.
A lot of validation/garbage collection in admin depends on updated_on so bumping it on reconcile seems correct.
May be we should revisit the approach commented here :

rill/runtime/config_reloader.go

Lines 28 to 32 in de14a79

	// to avoid repo handshake refresh on each reload we track last updatedon of each deployment
	// if the deployment has not changed skip the repo pull
	//
	// this can further be optimised to only check for properties that affect repo like url, branch etc but keeping it simple for now
	updatedOn map[string]time.Time