Andrey Chernyshev

I forgot to mention the exception does not happen if running under SBT (e.g. via sbt run). The app main has to be run from a command line - apparently...

Meanwhile, this is how async watch works for me (in case someone may still need it): ```python def get_range_end(prefix): return prefix[:-1] + chr(ord(prefix[-1]) + 1) async def on_event(self, key, value,...

Seems like adding `dynamic_startup_nodes=False` in cluster constructor solves the problem, at least for sync client.

Thanks very much for your prompt reply! I'll try with more debug level next time (don't want to cancel current job). I was running Scrapyd 1.5.0 in Kubernetes pod with...

[scrapyd1.zip](https://github.com/user-attachments/files/19402276/scrapyd1.zip) Here is the log from my scrapyd installation (used version from git master). It was able to run spider2 and spider3 concurrently for a while. Then notice a few...