Skip to content

Panic invalid memory address after reconnect a node #139

@davidroman0O

Description

@davidroman0O

Describe the bug
Using the example genstage, change it to have two nodes, connect one to the other, disconnect, reconnect and panic

It kills the producer process directly, the entire node does NOT crash

  • Open two terminals
  • First one: go run ./producer
  • Second one: go run ./consumer
  • They should communicate and trigger events
  • CTRL + C on consumer
  • Restart it again
  • The error should show up
2023/01/06 17:06:25 WARNING! Server terminated <336B493D.0.1011>["producer"]. Panic reason: "invalid memory address or nil 

To Reproduce
Steps to reproduce the behavior:

I've made a gist with all the code and a readme there https://gist.github.com/davidroman0O/db04fa224c710a4ae33d53d28304e88b

Expected behavior
Maybe a better error why it has a nil pointer? why? where? stacktrace?

Or maybe it should just not panic

Environment (please complete the following information):

  • OS: Tested on Windows 10 and QubeOS 4.11
  • Framework Version v1.999.220
  • Number of CPU 12

Additional context

I was just experimenting with the examples, trying to plug a bit of my codebase of some side projects and end up with this issue
I came back on the examples to test it again with very little modification and the error appear!

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions