Hello everyone, I am deploying Airbyte on a Google Compute Engine instance. The deployment is successful, and I'm able to run some synchronizations. However, in some cases during the syncs, it gets stuck in an infinite loop with the message

Pool queue size: 0, Active threads: 0

I noticed that the error occurs when there are several large tables selected, when only small tables are selected, it does not happen. There are even cases where a single table enters this loop. I have already updated the source and destination connector versions and completely reinstalled Airbyte, but the error persists. Current Configuration • Service: Google Compute Engine • Machine type: c2-standard-8 (8 vCPUs, 32 GB Memory) • Disk Size: 100GB • Airbyte Version: 2.0.19 Connector Versions: • source-bigquery: 0.4.4 • destination-mssql: 2.2.14 • destination-postgres: 2.4.7 Logs for BigQuery > MS SQL

```
Logs for BigQuery > Postgres
```

Based on the information in the provided knowledge sources, the issue you are experiencing—syncs getting stuck with the message `Pool queue size: 0, Active threads: 0`—has been reported in the context of Airbyte syncs, particularly with large tables or data volumes. From the logs and issue discussions, this behavior is observed when the destination (e.g., BigQuery) appears to process batches, but after a flush, the queue remains empty and no further progress is made. The logs repeatedly show:

[ASYNC WORKER INFO] Pool queue size: 0, Active threads: 0

and

[ASYNC QUEUE INFO] ... num records: 0, num bytes: 0 bytes, allocated bytes: 0 bytes ...

This state persists until the sync is manually cancelled. The logs also show that after flushing, the system sometimes receives a "null value for Job, nothing to waitFor" from BigQuery, which may indicate an issue with job submission or tracking in the destination connector:

WARN ...BigQueryUtils(waitForJobFinish):276 Received null value for Job, nothing to waitFor

Key Observations: - The issue is not resolved by updating connector versions or reinstalling Airbyte. - It is more likely to occur with large tables or when multiple large tables are selected. - The system appears to process and flush batches, but then enters a state where no new records are processed, and the worker threads are idle. Possible Causes (from the logs): - The destination connector (BigQuery in the referenced issue) may not be properly handling large batches or is failing silently after a flush, leading to a deadlock or idle state. - The "null value for Job" warning suggests that the BigQuery job may not be created or tracked correctly, which could prevent further progress. - There is no explicit error or out-of-memory message, and the memory usage appears within limits. What is not clear: - The knowledge sources do not provide a definitive root cause or a solution for this issue. - There is no mention of a workaround or configuration change that resolves the problem. Recommendation: - Since this appears to be a known but unresolved issue, you may want to follow or contribute to the ongoing discussion in the related GitHub issue for updates or potential fixes: [Sync hangs forever, saying nothing to do. When cancel sync, it says one message was there](https://github.com/airbytehq/airbyte/issues/58082). - If you have not already, consider providing detailed logs and configuration to the Airbyte team via GitHub to help diagnose the problem further. I'm sorry, but the knowledge sources do not contain enough information to provide a definitive solution to this issue. The problem seems to be under investigation by the Airbyte community and maintainers.