control forced checkpoint simulation via env var

apps/coordinator/src/index.ts

@@ -20,6 +20,10 @@ const HTTP_SERVER_PORT = Number(process.env.HTTP_SERVER_PORT || 8020);
 const NODE_NAME = process.env.NODE_NAME || "coordinator";
 const DEFAULT_RETRY_DELAY_THRESHOLD_IN_MS = 30_000;
 
+const FORCE_CHECKPOINT_SIMULATION = ["1", "true"].includes(
+  process.env.FORCE_CHECKPOINT_SIMULATION ?? "true"
+);
+
 const REGISTRY_HOST = process.env.REGISTRY_HOST || "localhost:5000";
 const CHECKPOINT_PATH = process.env.CHECKPOINT_PATH || "/checkpoints";
 const REGISTRY_TLS_VERIFY = process.env.REGISTRY_TLS_VERIFY === "false" ? "false" : "true";
@@ -319,7 +323,7 @@ class Checkpointer {
 
 class TaskCoordinator {
   #httpServer: ReturnType<typeof createServer>;
-  #checkpointer = new Checkpointer({ forceSimulate: true });
+  #checkpointer = new Checkpointer({ forceSimulate: FORCE_CHECKPOINT_SIMULATION });
 
   #prodWorkerNamespace: ZodNamespace<
     typeof ProdWorkerToCoordinatorMessages,

apps/docker-provider/src/index.ts

@@ -14,9 +14,14 @@ import { randomUUID } from "node:crypto";
 const MACHINE_NAME = process.env.MACHINE_NAME || "local";
 const COORDINATOR_PORT = process.env.COORDINATOR_PORT || 8020;
 const COORDINATOR_HOST = process.env.COORDINATOR_HOST || "127.0.0.1";
+
 const OTEL_EXPORTER_OTLP_ENDPOINT =
   process.env.OTEL_EXPORTER_OTLP_ENDPOINT || "http://0.0.0.0:4318";
 
+const FORCE_CHECKPOINT_SIMULATION = ["1", "true"].includes(
+  process.env.FORCE_CHECKPOINT_SIMULATION ?? "true"
+);
+
 const logger = new SimpleLogger(`[${MACHINE_NAME}]`);
 
 type InitializeReturn = {
@@ -293,7 +298,7 @@ class DockerTaskOperations implements TaskOperations {
 }
 
 const provider = new ProviderShell({
-  tasks: new DockerTaskOperations({ forceSimulate: true }),
+  tasks: new DockerTaskOperations({ forceSimulate: FORCE_CHECKPOINT_SIMULATION }),
   type: "docker",
 });