What causes "[Codex Batch Delete] 查询任务进度失败:" in jlcodes99/cockpit-tools?

No action needed for one-off occurrences — the loop retries automatically in 250ms.

[Codex Batch Delete] 查询任务进度失败:

Error message

[Codex Batch Delete] 查询任务进度失败:

What it means

Not a thrown error: a console.warn inside the Codex batch-delete polling loop in useCodexAccountsBaseController. Each 250ms pollJob tick calls the backend job-progress API inside try/catch; if that single query fails (e.g. transient IPC/backend error), the loop logs the warning and simply retries on the next tick. The batch delete continues unless the failure persists.

Source

Thrown at src/pages/useCodexAccountsBaseController.tsx:755

      if (!jobId || batchDeleteJob.status !== "running") return;
  
      let disposed = false;
      let timer: number | null = null;
      const pollJob = async () => {
        let shouldContinue = true;
        try {
          const nextJob = await codexService.getCodexBatchDelete(jobId);
          if (disposed) return;
          setBatchDeleteJob((current) =>
            current?.jobId === jobId ? nextJob : current,
          );
          if (nextJob.completed > batchDeleteRefreshedCompletedRef.current) {
            await refreshAccountsDuringBatchDelete();
            batchDeleteRefreshedCompletedRef.current = nextJob.completed;
          }
          shouldContinue = nextJob.status === "running";
        } catch (error) {
          console.warn("[Codex Batch Delete] 查询任务进度失败:", error);
        }
        if (!disposed && shouldContinue) {
          timer = window.setTimeout(pollJob, 250);
        }
      };
  
      timer = window.setTimeout(pollJob, 100);
      return () => {
        disposed = true;
        if (timer !== null) window.clearTimeout(timer);
      };
    }, [
      batchDeleteJob?.jobId,
      batchDeleteJob?.status,
      refreshAccountsDuringBatchDelete,
    ]);
  
    useEffect(() => {

View on GitHub (pinned to 1ed8b77992)

Solutions

No action needed for one-off occurrences — the loop retries automatically in 250ms.
If it fires repeatedly, verify the batch-delete backend task still exists and the progress query command is healthy.
Check backend logs at the same timestamp for the underlying cause (task not found, IPC failure).
Restart the app/backend process if progress queries persistently fail and the batch appears stuck.
Update to a build where progress-poll failures are surfaced in the batch-delete modal instead of only the console.

Defensive patterns

Strategy: try-catch

Try / catch

try {
  const nextJob = await queryBatchDeleteJobProgress(jobId);
  // proceed
} catch (error) {
  console.warn('[Codex Batch Delete] 查询任务进度失败:', error);
  // loop continues; consider a consecutive-failure counter to abort after N failures
}

Prevention

Track consecutive poll failures and surface an error in the UI after a threshold instead of silently retrying forever.
Add a timeout to the job-progress query so a hung backend doesn't stall the loop.
Ensure the backend job record outlives the polling session (don't GC while clients poll).

When it happens

Trigger: The pollJob timer fires and the promise resolving job progress (backend invoke / service call returning nextJob) rejects — backend process busy/restarting, IPC channel error, job already finished and cleaned up server-side.

Common situations: Deleting many Codex accounts in a batch while the backend is under load; rapid window operations suspending the webview timers; backend task record removed mid-poll so the progress query 404s/errors.

Related errors

AI-assisted analysis of jlcodes99/cockpit-tools@1ed8b77992 (2026-09-05). Data as JSON: /api/errors/b925c07c5f468016. Report an issue: GitHub.