Skip to content

fix(runners): resolve zombie-state and ensure consistent pool release #39

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
ttbombadil merged 4 commits into from
Mar 2, 2026
+135 −197
Merged

fix(runners): resolve zombie-state and ensure consistent pool release #39

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
1a1c99c
fix(runners): resolve zombie-state and ensure consistent pool release
ttbombadil Mar 2, 2026
a7ebe46
Merge branch 'main' into performance
ttbombadil Mar 2, 2026
20c98f4
fix(pool): prevent event listener leakage on runner recycle
ttbombadil Mar 2, 2026
c960c5a
test(pause-resume): make timing tests robust against listener cleanup
ttbombadil Mar 2, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
1 change: 0 additions & 1 deletion server/routes.ts
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -27,7 +27,6 @@ export async function registerRoutes(app: Express): Promise<Server> {
const logger = new Logger("Routes");
const httpServer = createServer(app);

// Initialize SandboxRunnerPool for managing runner instances
await initializeSandboxRunnerPool();

// Lightweight health endpoint for backend reachability checks
Expand Down
108 changes: 48 additions & 60 deletions server/routes/simulation.ws.ts
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -14,10 +14,11 @@ export type SimulationDeps = {
shouldSendSimulationEndMessage: (compileFailed: boolean) => boolean;
getLastCompiledCode: () => string | null;
logger: Logger;
runnerPool?: ReturnType<typeof getSandboxRunnerPool>;
};

// Return type exposes a small API used by other modules (test-reset)
export function registerSimulationWebSocket(httpServer: Server, deps: SimulationDeps & { runnerPool?: ReturnType<typeof getSandboxRunnerPool> }) {
export function registerSimulationWebSocket(httpServer: Server, deps: SimulationDeps) {
const { SandboxRunner, getSimulationRateLimiter, shouldSendSimulationEndMessage, getLastCompiledCode, logger, runnerPool } = deps;
const pool = runnerPool ?? getSandboxRunnerPool();

Expand Down Expand Up @@ -56,6 +57,32 @@ export function registerSimulationWebSocket(httpServer: Server, deps: Simulation
}
}

async function safeReleaseRunner(
state: { runner: InstanceType<typeof SandboxRunner> | null; isRunning: boolean; isPaused: boolean },
reason: string,
): Promise<void> {
if (!state.runner) {
return;
}

const runner = state.runner;
state.runner = null;
state.isRunning = false;
state.isPaused = false;

try {
await runner.stop();
} catch (error) {
logger.debug(`[SandboxRunnerPool] runner.stop() failed during ${reason}: ${error}`);
}

try {
await pool.releaseRunner(runner);
} catch (error) {
logger.warn(`[SandboxRunnerPool] releaseRunner failed during ${reason}: ${error}`);
}
}

wss.on("connection", (ws, req) => {
const url = req.url || "";
const urlParams = new URLSearchParams(url.split("?")[1] || "");
Expand Down Expand Up @@ -95,9 +122,7 @@ export function registerSimulationWebSocket(httpServer: Server, deps: Simulation

const clientState = clientRunners.get(ws);
if (clientState?.runner) {
clientState.runner.stop();
clientState.isRunning = false;
clientState.isPaused = false;
await safeReleaseRunner(clientState, "rate-limit");
}

sendMessageToClient(ws, {
Expand All @@ -114,10 +139,7 @@ export function registerSimulationWebSocket(httpServer: Server, deps: Simulation
const lastCompiledCode = getLastCompiledCode();
if (!lastCompiledCode) {
if (clientState.runner) {
await clientState.runner.stop();
// Release old runner back to pool
await pool.releaseRunner(clientState.runner);
clientState.runner = null;
await safeReleaseRunner(clientState, "missing-compiled-code");
}
clientState.isRunning = false;
clientState.isPaused = false;
Expand All @@ -127,26 +149,23 @@ export function registerSimulationWebSocket(httpServer: Server, deps: Simulation
break;
}

// Release old runner if exists
if (clientState.runner) {
await clientState.runner.stop();
await pool.releaseRunner(clientState.runner);
await safeReleaseRunner(clientState, "start-replace-existing");
}

// Acquire fresh runner from pool (not new instance)
try {
clientState.runner = await pool.acquireRunner();
logger.debug(`[SandboxRunnerPool] Acquired runner for client. Pool stats: ${JSON.stringify(pool.getStats())}`);
} catch (acquireError) {
logger.error(`[SandboxRunnerPool] Failed to acquire runner: ${acquireError}`);
} catch (error) {
logger.error(`[SandboxRunnerPool] Failed to acquire runner: ${error}`);
clientState.runner = null;
clientState.isRunning = false;
clientState.isPaused = false;
sendMessageToClient(ws, { type: "serial_output", data: "[ERR] Server overloaded. All runners busy. Please try again.\n" });
sendMessageToClient(ws, { type: "simulation_status", status: "stopped" });
break;
}

// Note: tempDir handling is already configured internally in SandboxRunner
clientState.isRunning = true;
clientState.isPaused = false;

Expand Down Expand Up @@ -177,19 +196,7 @@ export function registerSimulationWebSocket(httpServer: Server, deps: Simulation
try {
const cs = clientRunners.get(ws);
if (cs) {
cs.isRunning = false;
cs.isPaused = false;

// Release runner back to pool when simulation ends
if (cs.runner) {
try {
await pool.releaseRunner(cs.runner);
logger.debug(`[SandboxRunnerPool] Released runner on exit. Pool stats: ${JSON.stringify(pool.getStats())}`);
} catch (releaseErr) {
logger.warn(`[SandboxRunnerPool] Error releasing runner on exit: ${releaseErr}`);
}
cs.runner = null;
}
await safeReleaseRunner(cs, "onExit");
}

if (!shouldSendSimulationEndMessage(compileFailed)) return;
Expand All @@ -212,17 +219,10 @@ export function registerSimulationWebSocket(httpServer: Server, deps: Simulation
sendMessageToClient(ws, { type: "compilation_status", gccStatus: "error" });
sendMessageToClient(ws, { type: "simulation_status", status: "stopped" });
const cs = clientRunners.get(ws);
if (cs) {
cs.isRunning = false;
cs.isPaused = false;

// Release runner back to pool on compile error
if (cs.runner) {
pool.releaseRunner(cs.runner).catch(err => {
logger.warn(`[SandboxRunnerPool] Error releasing runner on compile error: ${err}`);
});
cs.runner = null;
}
if (cs) {
safeReleaseRunner(cs, "onCompileError").catch((error) => {
logger.warn(`[SandboxRunnerPool] safeReleaseRunner failed in onCompileError: ${error}`);
});
}
logger.error(`[Client Compile Error]: ${compileErr}`);
},
Expand Down Expand Up @@ -288,9 +288,7 @@ export function registerSimulationWebSocket(httpServer: Server, deps: Simulation
case "code_changed": {
const clientState = clientRunners.get(ws);
if (clientState?.runner && (clientState?.isRunning || clientState?.isPaused)) {
clientState.runner.stop();
clientState.isRunning = false;
clientState.isPaused = false;
await safeReleaseRunner(clientState, "code_changed");
sendMessageToClient(ws, { type: "simulation_status", status: "stopped" });
sendMessageToClient(ws, { type: "serial_output", data: "--- Simulation stopped due to code change ---\n" });
}
Expand All @@ -300,9 +298,7 @@ export function registerSimulationWebSocket(httpServer: Server, deps: Simulation
case "stop_simulation": {
const clientState = clientRunners.get(ws);
if (clientState?.runner) {
clientState.runner.stop();
clientState.isRunning = false;
clientState.isPaused = false;
await safeReleaseRunner(clientState, "stop_simulation");
}
sendMessageToClient(ws, { type: "simulation_status", status: "stopped" });
sendMessageToClient(ws, { type: "serial_output", data: "--- Simulation stopped ---\n" });
Expand Down Expand Up @@ -364,20 +360,19 @@ export function registerSimulationWebSocket(httpServer: Server, deps: Simulation
ws.on("close", async () => {
const clientState = clientRunners.get(ws);
if (clientState?.runner) {
await clientState.runner.stop();
// Release runner back to pool when client disconnects
await pool.releaseRunner(clientState.runner).catch(err => {
logger.warn(`[SandboxRunnerPool] Error releasing runner on client close: ${err}`);
});
clientState.runner = null;
await safeReleaseRunner(clientState, "ws-close");
}
clientRunners.delete(ws);
const rateLimiter = getSimulationRateLimiter();
rateLimiter.removeClient(ws);
logger.info(`Client disconnected. Remaining clients: ${wss.clients.size}`);
});

ws.on("error", (error) => {
ws.on("error", async (error) => {
const clientState = clientRunners.get(ws);
if (clientState?.runner) {
await safeReleaseRunner(clientState, "ws-error");
}
logger.error(`WebSocket error: ${error instanceof Error ? error.message : String(error)}`);
});
});
Expand All @@ -388,14 +383,7 @@ export function registerSimulationWebSocket(httpServer: Server, deps: Simulation

for (const [ws, clientState] of clientRunners.entries()) {
if (clientState.runner) {
try {
await clientState.runner.stop();
// Release runner back to pool during reset
await pool.releaseRunner(clientState.runner);
} catch (err) {
logger.debug(`Failed to stop/release runner during reset: ${err}`);
}
clientState.runner = null;
await safeReleaseRunner(clientState, "test-reset");
}
clientState.isRunning = false;
clientState.isPaused = false;
Expand Down
Loading
Loading