Scheduler and worker loops survive a database restart
Test and deploy / test (push) Successful in 1m0s
Test and deploy / deploy (push) Successful in 2m52s

index-scheduler's loop ended when Postgres restarted (the active-groups
query was outside its try), leaving live indexing stopped for ~3 hours
while the process stayed up. Each step's error is now caught and retried
next round; worker:spots and worker:deferred wait 30 s and claim again.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
This commit is contained in:
Daan MeijerandClaude Opus 5.5 committed 2026-10-07 13:02:08 +02:00
1 parent 61192690a7
commit e3e3ecc93d
3 files changed
+60 -41

No files matched your search

+22 -16
View File
@@ -31,23 +31,29 @@ export default class DeferredWorker extends BaseCommand {
let done = 0
try {
while (!Shutdown.requested && (!this.limit || done < this.limit)) {
const batch = await DeferredService.claim(Math.min(concurrency * 4, this.limit ? this.limit - done : Infinity))
if (batch.length === 0) {
if (this.limit) break
await Shutdown.sleep(IDLE_MS)
continue
try {
const batch = await DeferredService.claim(Math.min(concurrency * 4, this.limit ? this.limit - done : Infinity))
if (batch.length === 0) {
if (this.limit) break
await Shutdown.sleep(IDLE_MS)
continue
}
let next = 0
await Promise.all(
Array.from({ length: concurrency }, async () => {
while (next < batch.length) {
const status = await DeferredService.process(batch[next++])
counts[status] = (counts[status] ?? 0) + 1
}
})
)
done += batch.length
this.logger.info(`${done} deferred files processed: ${Object.entries(counts).map(([k, v]) => `${v} ${k}`).join(', ')}.`)
} catch (err: any) {
// E.g. Postgres restarting: wait and claim again, instead of ending the loop (LOG.md 2026-10-07 (3)).
this.logger.error(`Round failed, retrying in 30 s: ${err.message}`)
await Shutdown.sleep(30_000)
}
let next = 0
await Promise.all(
Array.from({ length: concurrency }, async () => {
while (next < batch.length) {
const status = await DeferredService.process(batch[next++])
counts[status] = (counts[status] ?? 0) + 1
}
})
)
done += batch.length
this.logger.info(`${done} deferred files processed: ${Object.entries(counts).map(([k, v]) => `${v} ${k}`).join(', ')}.`)
}
} finally {
await NntpService.shutdown()