Johan

A few hours later it crashed again. Including all log since last startup: ``` Nov 23 07:35:27 back-1 prometheus[71014]: ts=2022-11-23T06:35:27.914Z caller=main.go:543 level=info msg="Starting Prometheus Server" mode=server version="(version=2.39.1, branch=release-2.39, revision=0)" Nov...

Was this ever fixed? With latest release (minus the commit referenced in #534) and throwing lots of subscribres on single worker process, it does not seem to decrement properly once...

This was trigged with nchan 1.2.6 from 4a4e42967fd37c1cb57e2142632f6180dee49253, and nginx 1.17.6. I manage to repeat it every time. The actual output from above was from MacOS, but I've originally discovered...

Good question, I don't actually remember if and how.. Not even sure what I was using it for!

Hi, I have the same issue. 1.2.5 works, 1.2.6 crashes. Managed to reproduce it, with the following minimal test case: ``` import requests while True: try: print("Getting...") hdr = {...

Seems to be introduced by https://github.com/slact/nchan/commit/82b766cd133c060542c2420b19c13d74de6cc0c6, reverting that and it stops crashing.

For the record, this happens with Redis enabled too (previous test was without redis. Same config as below but remove redis stuff). nginx config used: ``` worker_processes 1; daemon off;...

Don't think I ever got it working, no. Ended up throwing it out and using an adapted version of http://code.activestate.com/recipes/578659-python-3-token-bucket/ instead