@genroc/eval-node 0.0.0-edge.bac4287 → 0.0.0-edge.c42aeea

package/dist/eval.js

@@ -20,13 +20,31 @@ class RealmFault extends Error {
         this.name = "RealmFault";
     }
 }
-export async function evaluate(req) {
+/** Thrown when `signal` aborts: the work was cancelled server-side, so there is no outcome to
+ *  report and worker.ts must release rather than answer. Distinct from RealmFault because the
+ *  runner did not fault -- nothing is wrong, the answer is simply no longer wanted. */
+export class Cancelled extends Error {
+    constructor() {
+        super("cancelled");
+        this.name = "Cancelled";
+    }
+}
+export async function evaluate(req, signal) {
     const budget = typeof req.timeout_ms === "number" ? req.timeout_ms : DEFAULT_TIMEOUT_MS;
     const worker = new Worker(REALM_URL);
     let timer;
+    let onAbort;
     try {
         return await new Promise((resolve, reject) => {
             timer = setTimeout(() => resolve(timedOut(budget)), budget);
+            // The abort is wired to the same promise as the budget, so both settle through the one
+            // finally below -- which is what guarantees the thread is gone before either returns.
+            if (signal) {
+                if (signal.aborted)
+                    reject(new Cancelled());
+                onAbort = () => reject(new Cancelled());
+                signal.addEventListener("abort", onAbort, { once: true });
+            }
             worker.once("message", (reply) => resolve(reply));
             // A script may end its own realm (`process.exit()`), which is not a throw and would
             // otherwise present as a hang until the budget expired. Our own terminate() raises
@@ -38,6 +56,8 @@ export async function evaluate(req) {
     }
     finally {
         clearTimeout(timer);
+        if (signal && onAbort)
+            signal.removeEventListener("abort", onAbort);
         // Awaited, and the whole point: on the timeout path a thread is still burning a core, and
         // resolving before it is gone would report an evaluation the machine is still running.
         await worker.terminate();

package/dist/worker.js

@@ -5,7 +5,7 @@
 //
 // See README.md for the contract, and specs/external-task-queue.md for the queue itself.
 import { readFileSync } from "node:fs";
-import { evaluate } from "./eval.js";
+import { Cancelled, evaluate } from "./eval.js";
 const SERVER = (process.env.GENROC_SERVER ?? "http://localhost:8448").replace(/\/$/, "");
 const WORKER_ID = process.env.WORKER_ID ?? `evaluator-${process.pid}`;
 // The credential, when the server runs with --auth token. A worker needs exactly the `worker`
@@ -185,7 +185,7 @@ async function answer(token, outcome) {
     console.error(`genroc refused the outcome for ${token}: ${JSON.stringify(data)}`);
     await release(token);
 }
-async function run(job) {
+async function run(job, signal) {
     let resolved;
     try {
         resolved = await resolveObjects(job);
@@ -206,9 +206,17 @@ async function run(job) {
     }
     let result;
     try {
-        result = await evaluate(req);
+        result = await evaluate(req, signal);
     }
     catch (err) {
+        // Cancelled is not a fault: the process was stopped while this ran. The claim still goes
+        // back — the row is terminal, so nothing re-claims it, and the release is what stops it
+        // waiting out a lease nobody is serving.
+        if (err instanceof Cancelled) {
+            console.log(`cancelled: ${job.token}`);
+            await release(job.token);
+            return;
+        }
         // The RUNNER faulted, not the script — the one class where a retry can help. There is no
         // error code for it on purpose: releasing the claim is how a queue spells "retryable", and
         // it puts the task in front of a different worker instead of burning the definition's
@@ -245,11 +253,22 @@ async function renewLoop() {
             tokens,
             lease_ms: LEASE_MS,
         });
-        // A short count means a claim lapsed and was taken over. Nothing to do about it — the work
-        // continues and its answer will be refused — but say so, because it is the signal that
-        // LEASE_MS is too short for what these scripts actually take.
-        if (ok && data?.renewed < tokens.length) {
-            console.error(`renewed ${data.renewed}/${tokens.length} claims; a lease lapsed under load`);
+        if (!ok)
+            continue;
+        // Cancelled: the process was stopped, so the evaluation is abandoned and the claim handed
+        // back. Aborting first is the point of the list — waiting for the script to finish would
+        // keep burning a core on work an operator has already stopped. The release rides run()'s
+        // own catch, so nothing is released twice.
+        for (const token of (data?.cancelled ?? [])) {
+            inFlight.get(token)?.abort.abort();
+        }
+        // Lost: someone else holds this claim now. The work continues and its answer will be
+        // refused — that is not fixable here — but say so, because it is the signal that LEASE_MS
+        // is too short for what these scripts actually take. Deliberately NOT released: the claim
+        // is the new holder's, and releasing would bump the epoch out from under it.
+        const lost = (data?.lost ?? []);
+        if (lost.length) {
+            console.error(`lost ${lost.length}/${tokens.length} claims; a lease lapsed under load`);
         }
     }
 }
@@ -258,8 +277,9 @@ async function pollLoop() {
         const free = CONCURRENCY - inFlight.size;
         const jobs = free > 0 ? await claim(free) : [];
         for (const job of jobs) {
-            inFlight.set(job.token, job);
-            void run(job).finally(() => inFlight.delete(job.token));
+            const abort = new AbortController();
+            inFlight.set(job.token, { job, abort });
+            void run(job, abort.signal).finally(() => inFlight.delete(job.token));
         }
         // Only idle when there was nothing to take: a full queue should be drained at the speed the
         // realms allow, not at the poll interval.

package/eval.ts

@@ -57,14 +57,32 @@ class RealmFault extends Error {
   }
 }
 
-export async function evaluate(req: EvalRequest): Promise<EvalResult> {
+/** Thrown when `signal` aborts: the work was cancelled server-side, so there is no outcome to
+ *  report and worker.ts must release rather than answer. Distinct from RealmFault because the
+ *  runner did not fault -- nothing is wrong, the answer is simply no longer wanted. */
+export class Cancelled extends Error {
+  constructor() {
+    super("cancelled");
+    this.name = "Cancelled";
+  }
+}
+
+export async function evaluate(req: EvalRequest, signal?: AbortSignal): Promise<EvalResult> {
   const budget = typeof req.timeout_ms === "number" ? req.timeout_ms : DEFAULT_TIMEOUT_MS;
 
   const worker = new Worker(REALM_URL);
   let timer: ReturnType<typeof setTimeout> | undefined;
+  let onAbort: (() => void) | undefined;
   try {
     return await new Promise<EvalResult>((resolve, reject) => {
       timer = setTimeout(() => resolve(timedOut(budget)), budget);
+      // The abort is wired to the same promise as the budget, so both settle through the one
+      // finally below -- which is what guarantees the thread is gone before either returns.
+      if (signal) {
+        if (signal.aborted) reject(new Cancelled());
+        onAbort = () => reject(new Cancelled());
+        signal.addEventListener("abort", onAbort, { once: true });
+      }
       worker.once("message", (reply: WorkerReply) => resolve(reply));
       // A script may end its own realm (`process.exit()`), which is not a throw and would
       // otherwise present as a hang until the budget expired. Our own terminate() raises
@@ -75,6 +93,7 @@ export async function evaluate(req: EvalRequest): Promise<EvalResult> {
     });
   } finally {
     clearTimeout(timer);
+    if (signal && onAbort) signal.removeEventListener("abort", onAbort);
     // Awaited, and the whole point: on the timeout path a thread is still burning a core, and
     // resolving before it is gone would report an evaluation the machine is still running.
     await worker.terminate();

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@genroc/eval-node",
-  "version": "0.0.0-edge.bac4287",
+  "version": "0.0.0-edge.c42aeea",
   "type": "module",
   "engines": {
     "node": ">=24"

package/worker.ts

@@ -6,7 +6,7 @@
 // See README.md for the contract, and specs/external-task-queue.md for the queue itself.
 
 import { readFileSync } from "node:fs";
-import { evaluate, type EvalRequest, type FailureKind } from "./eval.ts";
+import { Cancelled, evaluate, type EvalRequest, type FailureKind } from "./eval.ts";
 
 const SERVER = (process.env.GENROC_SERVER ?? "http://localhost:8448").replace(/\/$/, "");
 const WORKER_ID = process.env.WORKER_ID ?? `evaluator-${process.pid}`;
@@ -135,7 +135,12 @@ function asEvalRequest(input: unknown): EvalRequest | string {
   };
 }
 
-const inFlight = new Map<string, QueueTask>();
+/** A claim this worker is currently serving. The controller is how the renewal loop reaches
+ *  the evaluation: genroc cannot call us, so a cancellation arrives as an answer to our own
+ *  heartbeat and has to be delivered inward from there. */
+type Running = { job: QueueTask; abort: AbortController };
+
+const inFlight = new Map<string, Running>();
 let running = true;
 
 // Whether the last claim reached genroc. A worker polls several times a second, so an
@@ -203,7 +208,7 @@ async function answer(token: string, outcome: Record<string, unknown>): Promise<
   await release(token);
 }
 
-async function run(job: QueueTask): Promise<void> {
+async function run(job: QueueTask, signal: AbortSignal): Promise<void> {
   let resolved: unknown;
   try {
     resolved = await resolveObjects(job);
@@ -224,8 +229,16 @@ async function run(job: QueueTask): Promise<void> {
 
   let result;
   try {
-    result = await evaluate(req);
+    result = await evaluate(req, signal);
   } catch (err) {
+    // Cancelled is not a fault: the process was stopped while this ran. The claim still goes
+    // back — the row is terminal, so nothing re-claims it, and the release is what stops it
+    // waiting out a lease nobody is serving.
+    if (err instanceof Cancelled) {
+      console.log(`cancelled: ${job.token}`);
+      await release(job.token);
+      return;
+    }
     // The RUNNER faulted, not the script — the one class where a retry can help. There is no
     // error code for it on purpose: releasing the claim is how a queue spells "retryable", and
     // it puts the task in front of a different worker instead of burning the definition's
@@ -263,11 +276,22 @@ async function renewLoop(): Promise<void> {
       tokens,
       lease_ms: LEASE_MS,
     });
-    // A short count means a claim lapsed and was taken over. Nothing to do about it — the work
-    // continues and its answer will be refused — but say so, because it is the signal that
-    // LEASE_MS is too short for what these scripts actually take.
-    if (ok && data?.renewed < tokens.length) {
-      console.error(`renewed ${data.renewed}/${tokens.length} claims; a lease lapsed under load`);
+    if (!ok) continue;
+
+    // Cancelled: the process was stopped, so the evaluation is abandoned and the claim handed
+    // back. Aborting first is the point of the list — waiting for the script to finish would
+    // keep burning a core on work an operator has already stopped. The release rides run()'s
+    // own catch, so nothing is released twice.
+    for (const token of (data?.cancelled ?? []) as string[]) {
+      inFlight.get(token)?.abort.abort();
+    }
+    // Lost: someone else holds this claim now. The work continues and its answer will be
+    // refused — that is not fixable here — but say so, because it is the signal that LEASE_MS
+    // is too short for what these scripts actually take. Deliberately NOT released: the claim
+    // is the new holder's, and releasing would bump the epoch out from under it.
+    const lost = (data?.lost ?? []) as string[];
+    if (lost.length) {
+      console.error(`lost ${lost.length}/${tokens.length} claims; a lease lapsed under load`);
     }
   }
 }
@@ -277,8 +301,9 @@ async function pollLoop(): Promise<void> {
     const free = CONCURRENCY - inFlight.size;
     const jobs = free > 0 ? await claim(free) : [];
     for (const job of jobs) {
-      inFlight.set(job.token, job);
-      void run(job).finally(() => inFlight.delete(job.token));
+      const abort = new AbortController();
+      inFlight.set(job.token, { job, abort });
+      void run(job, abort.signal).finally(() => inFlight.delete(job.token));
     }
     // Only idle when there was nothing to take: a full queue should be drained at the speed the
     // realms allow, not at the poll interval.