CDC Driver creates too many threads

[nidyaonur]

We've integrated this CDC driver into several microservices within our system. During profiling, we observed that the driver spawns approximately 8000 threads, which seems excessive. Upon further investigation, it appears that the driver is not grouping vnodes efficiently, instead creating a separate thread group per vnode—and then multiplying that count by the number of monitored tables.

Is this the intended behavior, or is there a configuration or optimization we're missing?

While this doesn't currently cause performance degradation—since most of these threads remain idle—it does introduce significant noise in tracing and profiling, making it difficult to monitor our services effectively.

Bonus Question:
Is this repository still actively maintained and recommended for production use?

[sylwiaszunejko]

Is this repository still actively maintained and recommended for production use?

@dkropachev I don't think so, but maybe you could correct me

[dkropachev]

Is this repository still actively maintained and recommended for production use?

@dkropachev I don't think so, but maybe you could correct me

Unfortunately not, but we need to revive it.

[nidyaonur]

Unfortunately not, but we need to revive it.

@dkropachev I have recently spent a lot of time using and debugging it. Let me know how I can contribute.