prestodb/presto · error · PrestoException

SERVER_SHUTTING_DOWN

SERVER_SHUTTING_DOWN

Error message

Server is shutting down

What it means

Thrown by HiveSplitManager's RejectedExecutionHandler when a task is submitted to the split-loading executor after the Presto server has begun shutting down and its thread pools reject new work. It signals the query cannot proceed because the coordinator/worker is going away.

Source

Thrown at presto-hive/src/main/java/com/facebook/presto/hive/HiveSplitManager.java:1122

    private static class ErrorCodedExecutor
            implements Executor
    {
        private final Executor delegate;

        private ErrorCodedExecutor(Executor delegate)
        {
            this.delegate = requireNonNull(delegate, "delegate is null");
        }

        @Override
        public void execute(Runnable command)
        {
            try {
                delegate.execute(command);
            }
            catch (RejectedExecutionException e) {
                throw new PrestoException(SERVER_SHUTTING_DOWN, "Server is shutting down", e);
            }
        }
    }
}

View on GitHub (pinned to 55bb57d202)

Solutions

  1. Wait for the server restart to complete and re-run the query.
  2. Enable graceful shutdown and drain active queries before stopping (e.g., shutdown timeout so queries finish).
  3. Use a coordinator with query retry/ha proxy so clients resubmit during rolling restarts.
Defensive patterns

Strategy: retry

Try / catch

try { runQuery(sql); } catch (PrestoException e) {
    if (SERVER_SHUTTING_DOWN.toErrorCode().getCode() == e.getErrorCode().getCode()) {
        retryWithBackoff(sql); // resubmit after server restart completes
    } else { throw e; }
}

Prevention

When it happens

Trigger: Split loading tasks submitted via executor.execute() while the server is stopping; RejectedExecutionException from the underlying delegate executor is wrapped into SERVER_SHUTTING_DOWN.

Common situations: Deploying a new version or restarting Presto while queries run; server crash or SIGTERM during query execution; systemctl stop / graceful shutdown racing active queries.

Related errors

AI-assisted analysis of prestodb/presto@55bb57d202 (2026-09-04). Data as JSON: /api/errors/730810ae0d60937c. Report an issue: GitHub.