v0.13.0

.github/workflows/release.yml

@@ -93,7 +93,7 @@ jobs:
           password: ${{ secrets.GITHUB_TOKEN }}
 
       - name: Build and push
-        uses: docker/build-push-action@v6
+        uses: docker/build-push-action@v7
         with:
           context: web
           push: true
@@ -123,7 +123,7 @@ jobs:
           password: ${{ secrets.GITHUB_TOKEN }}
 
       - name: Build and push
-        uses: docker/build-push-action@v6
+        uses: docker/build-push-action@v7
         with:
           context: registry
           push: true

agent/cmd/agent/main.go

@@ -4,6 +4,7 @@ import (
 	"context"
 	"flag"
 	"log"
+	"net/url"
 	"os"
 	"os/signal"
 	"path/filepath"
@@ -33,6 +34,21 @@ var (
 	dataDir    string = paths.DataDir
 )
 
+func sanitizedEndpoint(endpoint string) string {
+	parsedURL, err := url.Parse(endpoint)
+	if err != nil || parsedURL.User == nil {
+		return endpoint
+	}
+
+	if _, hasPassword := parsedURL.User.Password(); hasPassword {
+		parsedURL.User = url.UserPassword("xxxxx", "xxxxx")
+	} else {
+		parsedURL.User = url.User("xxxxx")
+	}
+
+	return parsedURL.String()
+}
+
 func main() {
 	var (
 		controlPlaneURL     string
@@ -274,7 +290,7 @@ func main() {
 	var metricsSender agent.MetricsSender
 
 	if logsEndpoint != "" {
-		log.Println("[logs] log collection enabled, endpoint:", logsEndpoint)
+		log.Println("[logs] log collection enabled, endpoint:", sanitizedEndpoint(logsEndpoint))
 		logsSender = logs.NewVictoriaLogsSender(logsEndpoint, config.ServerID)
 		logCollector = logs.NewCollector(logsSender, dataDir)
 		if isProxy {
@@ -287,7 +303,7 @@ func main() {
 	}
 
 	if metricsEndpoint != "" {
-		log.Println("[metrics] metrics collection enabled, endpoint:", metricsEndpoint)
+		log.Println("[metrics] metrics collection enabled, endpoint:", sanitizedEndpoint(metricsEndpoint))
 		metricsSender = metrics.NewVictoriaMetricsSender(metricsEndpoint, config.ServerID)
 	}
 

deployment/README.md

@@ -68,7 +68,7 @@ start, so scaling `WEB_REPLICAS` does not run migrations from every replica.
 
 ## Database Migrations
 
-Schema is synced automatically by the one-shot `migrate` service via `drizzle-kit push`. This approach auto-confirms non-destructive changes (adding tables, columns, indexes) but will **not** auto-apply destructive changes like dropping columns or tables — those require manual intervention. If schema sync fails, `web` startup is blocked; inspect the failure with `docker compose -f compose.production.yml logs migrate`.
+Schema is synced automatically by the one-shot `migrate` service via `drizzle-kit push --force`. This keeps deployment non-interactive, including schema changes Drizzle classifies as data-loss operations such as dropping columns. If schema sync fails, `web` startup is blocked; inspect the failure with `docker compose -f compose.production.yml logs migrate`.
 
 **Future plan:** Once the schema stabilizes, switch to `drizzle-kit generate` + `drizzle-orm migrate()` with pre-generated SQL migration files. This will eliminate the esbuild/drizzle-kit dependency from the production image.
 

deployment/compose.postgres.yml

@@ -90,7 +90,7 @@ services:
     depends_on:
       postgres:
         condition: service_healthy
-    command: ["sh", "-c", "echo y | npx drizzle-kit push"]
+    command: ["npx", "drizzle-kit", "push", "--force"]
     restart: on-failure
 
   web:

deployment/compose.production.yml

@@ -69,7 +69,7 @@ services:
       - INNGEST_SIGNING_KEY=${INNGEST_SIGNING_KEY}
       - INNGEST_EVENT_KEY=${INNGEST_EVENT_KEY}
       - ALLOW_SIGNUP=${ALLOW_SIGNUP:-false}
-    command: ["sh", "-c", "echo y | npx drizzle-kit push"]
+    command: ["npx", "drizzle-kit", "push", "--force"]
     restart: on-failure
 
   web:

docs/services/scaling.mdx

@@ -5,15 +5,13 @@ description: "Replicas, placement, and server pinning."
 
 ## Replicas
 
-Each service can run multiple replicas across your cluster. Set the replica count from the service settings — the platform distributes containers across available servers.
+Each service can run multiple replicas across your cluster. Configure how many replicas run on each server from the service settings.
 
 Replica count ranges from 1 to 10 per service.
 
-## Auto-Placement
+## Placement
 
-By default, **auto-placement** is enabled. The control plane decides which servers run each replica based on available capacity.
-
-When auto-placement is disabled, you manually configure how many replicas run on each server using per-server replica assignments.
+Services use manual placement. Select the target servers and replica counts explicitly before deploying.
 
 ## Server Pinning
 
@@ -26,6 +24,6 @@ You can also manually lock any service to a specific server by setting the locke
 ## Limitations
 
 - Stateful services are limited to 1 replica.
-- Stateful services cannot use auto-placement — they are always pinned to their locked server.
+- Stateful services are always pinned to their locked server.
 - Stateful services do not automatically fail over to another server.
 - Maximum 10 replicas per service.

docs/services/volumes.mdx

@@ -44,7 +44,7 @@ You can restore a volume from any completed backup. The restore process download
 
 ## Limitations
 
-- Services with volumes are locked to a single server — they cannot be auto-placed across multiple nodes.
+- Services with volumes are locked to a single server.
 - Replica count is fixed at 1 for stateful services.
 - Volume data lives on the host filesystem and is not replicated to other servers.
 - If the server is lost, data is only recoverable from completed backups.

web/actions/projects.ts

@@ -31,16 +31,11 @@ import {
 import { DEFAULT_RESOURCE_LIMITS } from "@/lib/constants";
 import { inngest } from "@/lib/inngest/client";
 import { inngestEvents } from "@/lib/inngest/events";
-import {
-	calculateResourceAwarePlacement,
-	replaceServiceReplicaPlacements,
-} from "@/lib/placement";
 import { allocatePort } from "@/lib/port-allocation";
 import {
 	containerPathSchema,
 	githubRepoUrlSchema,
 	nameSchema,
-	replicaCountSchema,
 	volumeNameSchema,
 } from "@/lib/schemas";
 import type {
@@ -465,7 +460,6 @@ export async function createService(input: CreateServiceInput) {
 		githubRootDir,
 		replicas: 1,
 		stateful: false,
-		autoPlace: true,
 		resourceCpuLimit: resourceLimits.cpuCores,
 		resourceMemoryLimitMb: resourceLimits.memoryMb,
 	});
@@ -732,14 +726,9 @@ export async function restoreDeletedService(serviceId: string) {
 		if (activeReplica?.serverStatus === "online") {
 			targetServerId = activeReplica.serverId;
 		} else {
-			const placements = await calculateResourceAwarePlacement(service, 1);
-			const targetPlacement = placements[0];
-			if (!targetPlacement) {
-				throw new Error("Cannot restore because no online server is available");
-			}
-
-			targetServerId = targetPlacement.serverId;
-			await replaceServiceReplicaPlacements(serviceId, placements);
+			throw new Error(
+				"Cannot restore because the selected server is unavailable",
+			);
 		}
 	}
 
@@ -1352,14 +1341,14 @@ export async function addServiceVolume(
 			throw new Error("Service not found");
 		}
 
-		let totalReplicas = service.replicas;
-		if (!service.autoPlace) {
-			const configuredReplicas = await db
-				.select({ count: serviceReplicas.count })
-				.from(serviceReplicas)
-				.where(eq(serviceReplicas.serviceId, serviceId));
-			totalReplicas = configuredReplicas.reduce((sum, r) => sum + r.count, 0);
-		}
+		const configuredReplicas = await db
+			.select({ count: serviceReplicas.count })
+			.from(serviceReplicas)
+			.where(eq(serviceReplicas.serviceId, serviceId));
+		const totalReplicas = configuredReplicas.reduce(
+			(sum, r) => sum + r.count,
+			0,
+		);
 
 		if (totalReplicas > 1) {
 			throw new Error(
@@ -1391,7 +1380,7 @@ export async function addServiceVolume(
 		if (!service.stateful) {
 			await db
 				.update(services)
-				.set({ stateful: true, autoPlace: false })
+				.set({ stateful: true })
 				.where(eq(services.id, serviceId));
 		}
 
@@ -1450,74 +1439,13 @@ export async function removeServiceVolume(volumeId: string) {
 	if (remainingVolumes.length === 0 && service.stateful) {
 		await db
 			.update(services)
-			.set({ stateful: false, autoPlace: true })
+			.set({ stateful: false })
 			.where(eq(services.id, service.id));
 	}
 
 	return { success: true };
 }
 
-export async function updateServiceAutoPlace(
-	serviceId: string,
-	autoPlace: boolean,
-) {
-	const service = await getService(serviceId);
-	if (!service) {
-		throw new Error("Service not found");
-	}
-
-	if (service.stateful && autoPlace) {
-		throw new Error(
-			"Services with volumes cannot use auto-placement. Remove volumes first.",
-		);
-	}
-
-	await db
-		.update(services)
-		.set({ autoPlace })
-		.where(eq(services.id, serviceId));
-
-	if (autoPlace) {
-		await db
-			.delete(serviceReplicas)
-			.where(eq(serviceReplicas.serviceId, serviceId));
-	}
-
-	return { success: true };
-}
-
-export async function updateServiceReplicas(
-	serviceId: string,
-	replicas: number,
-) {
-	try {
-		const validatedReplicas = replicaCountSchema.parse(replicas);
-
-		const service = await getService(serviceId);
-		if (!service) {
-			throw new Error("Service not found");
-		}
-
-		if (service.stateful && validatedReplicas > 1) {
-			throw new Error(
-				"Services with volumes can only have 1 replica. Remove volumes first to scale up.",
-			);
-		}
-
-		await db
-			.update(services)
-			.set({ replicas: validatedReplicas })
-			.where(eq(services.id, serviceId));
-
-		return { success: true };
-	} catch (error) {
-		if (error instanceof ZodError) {
-			throw new Error(getZodErrorMessage(error, "Invalid replica count"));
-		}
-		throw error;
-	}
-}
-
 export async function updateServiceBackupSettings(
 	serviceId: string,
 	backupEnabled: boolean,

web/actions/settings.ts

@@ -1,32 +1,23 @@
 "use server";
 
-import { setSetting } from "@/db/queries";
 import { revalidatePath } from "next/cache";
 import isEmail from "validator/es/lib/isEmail";
+import { ZodError } from "zod";
+import { setSetting } from "@/db/queries";
+import { buildTimeoutSchema } from "@/lib/schemas";
 import {
-	SETTING_KEYS,
 	type EmailAlertsConfig,
 	emailAlertsConfigSchema,
+	SETTING_KEYS,
 } from "@/lib/settings-keys";
-import { ZodError } from "zod";
 import { getZodErrorMessage } from "@/lib/utils";
-import { buildTimeoutSchema } from "@/lib/schemas";
 
 export async function updateBuildServers(serverIds: string[]) {
 	await setSetting(SETTING_KEYS.SERVERS_ALLOWED_FOR_BUILDS, serverIds);
 	revalidatePath("/dashboard/settings");
 	return { success: true };
 }
 
-export async function updateExcludedServers(serverIds: string[]) {
-	await setSetting(
-		SETTING_KEYS.SERVERS_EXCLUDED_FROM_WORKLOAD_PLACEMENT,
-		serverIds,
-	);
-	revalidatePath("/dashboard/settings");
-	return { success: true };
-}
-
 export async function updateBuildTimeout(minutes: number) {
 	try {
 		const validatedMinutes = buildTimeoutSchema.parse(minutes);

web/app/api/servers/route.ts

@@ -1,14 +1,11 @@
 export const dynamic = "force-dynamic";
 
-import { auth } from "@/lib/auth";
 import { headers } from "next/headers";
 import { db } from "@/db";
 import { servers } from "@/db/schema";
-import { notInArray } from "drizzle-orm";
-import { getSetting } from "@/db/queries";
-import { SETTING_KEYS } from "@/lib/settings-keys";
+import { auth } from "@/lib/auth";
 
-export async function GET(request: Request) {
+export async function GET() {
 	const session = await auth.api.getSession({
 		headers: await headers(),
 	});
@@ -17,18 +14,7 @@ export async function GET(request: Request) {
 		return Response.json({ error: "Unauthorized" }, { status: 401 });
 	}
 
-	const url = new URL(request.url);
-	const forPlacement = url.searchParams.get("forPlacement") === "true";
-
-	let excludedIds: string[] = [];
-	if (forPlacement) {
-		excludedIds =
-			(await getSetting<string[]>(
-				SETTING_KEYS.SERVERS_EXCLUDED_FROM_WORKLOAD_PLACEMENT,
-			)) || [];
-	}
-
-	const query = db
+	const data = await db
 		.select({
 			id: servers.id,
 			name: servers.name,
@@ -41,14 +27,8 @@ export async function GET(request: Request) {
 			resourcesDisk: servers.resourcesDisk,
 			meta: servers.meta,
 		})
-		.from(servers);
-
-	const data =
-		excludedIds.length > 0
-			? await query
-					.where(notInArray(servers.id, excludedIds))
-					.orderBy(servers.createdAt)
-			: await query.orderBy(servers.createdAt);
+		.from(servers)
+		.orderBy(servers.createdAt);
 
 	return Response.json(data);
 }