Hey guys. Tried using Rubinius to improve our Sidekiq processing.
@tmornini I commented on the gist, but I'll recomment here:
The memory violation appears to happen inside yajl doing a malloc call. Until the EM issues is addressed, it's not easy to determine whether yajl got bad data or if there's some issue with Rubinius C-API that yajl is using.
Looks like this project shows some issues, also #2102. Do you think there's a way we could run those specs and this code to make sure it all works well?
Sorry for the delay, have been seriously head-down.
I'll make the code available in a few of weeks. :-)
Did you get to setting this up somewhere? We've also fixed some issues that showed with using Sidekiq, if you could try again to see if there are still issues that would be great!
Does this issue still happen? We've fixed a bunch of additional issues that showed up with Sidekiq that could result in crashes.
Unknown. We're no longer using Sidekiq. Feel free to close.
Ok, closing this one then. If you encounter something with a different sidekiq app then, please open a new issue.
@tmornini out of curiosity, did you switch to something instead of Sidekiq or are you no longer using that functionality?