apache/pulsar · warning · java.lang.RuntimeException

${e}

Error message

${e}

What it means

shutdown() performs a forced, non-graceful teardown by delegating to the v4 client's shutdown(); if that throws org.apache.pulsar.client.api.PulsarClientException, it is wrapped in a plain java.lang.RuntimeException whose message is the exception's toString (the class name plus message). This signals the shutdown path itself failed and the JVM-level cleanup did not complete cleanly.

Source

Thrown at pulsar-client-v5/src/main/java/org/apache/pulsar/client/impl/v5/PulsarClientV5.java:122

            v4Client.close();
        } catch (org.apache.pulsar.client.api.PulsarClientException e) {
            throw new PulsarClientException(e.getMessage(), e);
        }
    }

    @Override
    public CompletableFuture<Void> closeAsync() {
        return v4Client.closeAsync().exceptionally(ex -> {
            throw new CompletionException(new PulsarClientException(ex.getMessage(), ex));
        });
    }

    @Override
    public void shutdown() {
        try {
            v4Client.shutdown();
        } catch (org.apache.pulsar.client.api.PulsarClientException e) {
            throw new RuntimeException(e);
        }
    }
}

View on GitHub (pinned to 820761864e)

Solutions

1. Read the wrapped RuntimeException's cause (the PulsarClientException) for the root failure
2. Reorder lifecycle: call shutdown() only once and only after producers/consumers are stopped
3. If the cause is benign (already closed), guard the call with a closed flag or catch-and-log

Example fix

// before
client.shutdown(); // raw RuntimeException on failure
// after
try {
    client.shutdown();
} catch (RuntimeException e) {
    log.warn("Shutdown failed", e.getCause());
}

if (clientShutdown) return; // guard with your own lifecycle flag

try {
    client.shutdown();
} catch (RuntimeException e) {
    log.warn("Shutdown failed: {}", e.getCause());
}

When it happens

Trigger: Calling PulsarClientV5.shutdown() when the underlying v4 client's shutdown throws — e.g. I/O failures closing connections, or internal state indicating the client is already shut down.

Common situations: Shutdown hooks / JVM exit paths calling shutdown() during a network outage; calling shutdown() after close() (double teardown); container termination sequences where the broker is already gone.

Related errors

• Executor %s should have been shutdown before entering the te
• ProducerCache is already closed
• Timeout during managed ledger close
• ManagedLedgerFactory is already closed.
• ManagedLedger ${name} has already been closed

AI-assisted analysis of apache/pulsar@820761864e (2026-09-06). Data as JSON: /api/errors/5e048d83cd9854d4. Report an issue: GitHub.