Node.js Cloud Functions subscribe to a Pub/Sub topic and write to BigQuery. The topic's message processing rate is far higher than expected, but no errors appear in Cloud Logging. What are two likely causes? (Choose two.)

The subscriber code cannot keep up with the messages because Cloud Functions have resource limitations (CPU, memory, concurrency). If the message processing rate exceeds these limits, messages will back up in Pub/Sub, but no errors will be logged by the function itself if it's simply overwhelmed rather than failing. The subscriber code does not acknowledge messages, which means Pub/Sub will repeatedly redeliver the unacknowledged messages, further exacerbating the processing backlog. This creates a cycle where messages are processed slowly (or not at all) and then redelivered, leading to a high message processing rate in Pub/Sub without corresponding successful BigQuery writes or logged errors. Publisher throughput quota being too small would prevent messages from being published in the first place, not cause a high processing rate. Total outstanding messages exceeding 10 MB is a limit for a single pull request, not a general backlog issue. Error handling not properly catching run-time errors would result in errors appearing in Cloud Logging.