Rebuild: the coordinator stops on SIGTERM

It waited for every chunk in flight, and a connection the server held
back (after killed coordinators left sessions open) waits up to the
10-minute reply timeout: it never seemed to stop. It now waits at most
15 s, then closes all connections, busy ones too; the greeting and login
time out after 30 s. And run() ends with terminate(), as edge:node does.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

app/services/nntp_pool.ts

@@ -10,6 +10,7 @@ import { XoverClient } from '#lib/nntp/index'
  */
 class NntpPool {
   private idle: XoverClient[] = []
+  private busy = new Set<XoverClient>()
   private open = 0
   private waiting: ((client: XoverClient) => void)[] = []
 
@@ -19,6 +20,7 @@ class NntpPool {
 
   async use<T>(work: (client: XoverClient) => Promise<T>): Promise<T> {
     const client = await this.acquire()
+    this.busy.add(client)
     let broken = false
     try {
       return await work(client)
@@ -26,6 +28,7 @@ class NntpPool {
       broken = true
       throw error
     } finally {
+      this.busy.delete(client)
       this.release(client, broken)
     }
   }
@@ -71,8 +74,10 @@ class NntpPool {
     else this.idle.push(client)
   }
 
+  /** Closes every connection, also those in use: their commands fail. */
   async close() {
-    for (const client of this.idle.splice(0)) client.close()
+    for (const client of [...this.idle.splice(0), ...this.busy]) client.close()
+    this.busy.clear()
     this.open = 0
   }
 }

commands/coordinator_run.ts

@@ -11,6 +11,7 @@ const TICK_MS = 2000
 const PLAN_EVERY_MS = 60 * 60 * 1000
 const MERGE_EVERY_MS = 60 * 1000
 const MERGE_BUDGET_MS = 30 * 1000
+const STOP_WAIT_MS = 15 * 1000
 
 /**
  * The coordinator's background work (REBUILD.md, part 1): the prefetcher,
@@ -30,6 +31,7 @@ export default class CoordinatorRun extends BaseCommand {
     const done = new Promise<void>((resolve) => (stopped = resolve))
     // bin/console turns SIGTERM into app.terminate: wait for the spooling in flight first.
     this.app.terminating(async () => {
+      this.logger.info('Stopping: waiting for the chunks in flight.')
       stopping = true
       prefetcher.stop()
       await done
@@ -83,8 +85,13 @@ export default class CoordinatorRun extends BaseCommand {
       await sleep(TICK_MS)
     }
     await prefetching
-    while (prefetcher.active > 0) await sleep(200)
+    // The chunks in flight, for at most STOP_WAIT_MS (a connection the server holds back can take minutes); then closed.
+    const waitUntil = Date.now() + STOP_WAIT_MS
+    while (prefetcher.active > 0 && Date.now() < waitUntil) await sleep(200)
+    if (prefetcher.active > 0) this.logger.info(`Closing ${prefetcher.active} connections still fetching.`)
     await NntpPool.close()
+    this.logger.info('Coordinator stopped.')
     stopped!()
+    await this.terminate()
   }
 }

lib/nntp/xover_client.ts

@@ -67,7 +67,8 @@ export class XoverClient {
       ? tls.connect({ host: options.host, port: options.port, servername: options.servername || options.host })
       : net.connect({ host: options.host, port: options.port })
     const client = new XoverClient(sock, options.timeoutMs ?? 120_000)
-    sock.setTimeout(client.timeoutMs)
+    // The greeting and login are short: a server holding a connection back (too many open) fails it in 30 s, not timeoutMs.
+    sock.setTimeout(Math.min(client.timeoutMs, 30_000))
     try {
       await client.expect(await client.line(), ['200', '201'])
       if (options.user) {