Valery Piashchynski

@Smolevich Friendly ping :)

@Smolevich could you please also attach debug logs?

Hey @vokybykov 👋🏻 Unfortunately, these are regular error logs, not debug. You need to update the log level in the configuration to debug.

@Smolevich Hard to say, depends on IO on the particular pod/vm. But as far as I see, your log contains many `no free workers in the pool` messages. When RR...

@vokybykov Could you please send logs to my email in my GitHub profile?

Hey @Smolevich 👋🏻 Could you please create a repository with the sample to reproduce this problem (worker+starter)? This will help me to instantiate a test suite and quickly find a...

Hey @Smolevich 👋🏻 Thank you for the report. I'll check it. Please note that this feature is in beta and not supposed to be in production.

I'll get into this problem next week after multiple pool support for the jobs plugin. Since this is a highly concurrent code (auto-scaler), I will add more high-intensity testing, especially...

Hey @Smolevich 👋🏻 Responding to your questions: 1. All workers, before being available to execute any request, receive a wait4 system call to prevent being zombie after SIGKILL (or any...

@Smolevich I've opened a PR with a dynamic allocator rework, but don't have time to test, sorry. If that is ok, and you at least know Go a little, could...